Data sources

Where every number comes from, under what licence, and what is still unresolved.

Data is retrieved with update.py, validated, and committed before the site is built. The live site never downloads benchmark data. Download the complete provenance table (CSV).

Epoch AI Benchmark Hub

Publisher
Epoch AI
Licence
CC-BY-4.0 (terms). Stated in the README inside benchmark_data.zip: free to use, distribute and reproduce with credit to the source and authors under CC BY. Checked 2026-10-28.
Attribution
Epoch AI, 'Capabilities & benchmarking'. Published online at epoch.ai. Retrieved from https://epoch.ai/benchmarks (CC BY 4.0).
Benchmarks extracted
FrontierMath Tiers 1-3 (v2), OTIS Mock AIME 2024-2025, GPQA Diamond, SimpleQA Verified, SWE-bench Verified (Epoch runs), FrontierCode (reported by Cognition)
Last data check
29 Oct 2026, 14:36 UTC
Last verified source update
Newest run start date among used results: 13 Oct 2026
Filtering and exclusions
Six benchmark files only. Rows are dropped if any source field mentions Artificial Analysis. Rows matched by exact 'Model version' to config/models.json. The score column named in the benchmark metadata is used.
Published here
Yes

Files used

Files used from Epoch AI Benchmark Hub
FileLocationSHA-256 (start)Content first seen
benchmark_metadatahttps://epoch.ai/data/benchmark_data.zip (member: benchmark_metadata.csv)979c73b823a2ea9f…28 Oct 2026
frontiercode_externalhttps://epoch.ai/data/benchmark_data.zip (member: frontiercode_external.csv)0ad9faa3508b3d42…29 Oct 2026
frontiermath_tiers_1_3_v2https://epoch.ai/data/benchmark_data.zip (member: frontiermath_tiers_1_3_v2.csv)c73649bc6e47b6e4…28 Oct 2026
gpqa_diamondhttps://epoch.ai/data/benchmark_data.zip (member: gpqa_diamond.csv)cf359c80bae654cc…29 Oct 2026
model_metadatahttps://epoch.ai/data/benchmark_data.zip (member: model_metadata.csv)fcc3e9dc3430e813…29 Oct 2026
otis_mock_aime_2024_2025https://epoch.ai/data/benchmark_data.zip (member: otis_mock_aime_2024_2025.csv)79534a01d4ab97ec…28 Oct 2026
simpleqa_verifiedhttps://epoch.ai/data/benchmark_data.zip (member: simpleqa_verified.csv)2dc690b3143b5098…28 Oct 2026
swe_bench_verifiedhttps://epoch.ai/data/benchmark_data.zip (member: swe_bench_verified.csv)6480176a5d9b1573…28 Oct 2026

Unresolved questions

  • The five benchmark files Epoch runs itself (FrontierMath Tiers 1-3, OTIS Mock AIME, GPQA Diamond, SimpleQA Verified, SWE-bench Verified) contain no explicit row-level provenance field. We exclude any row whose source fields mention Artificial Analysis, and found none in the files we use, but cannot positively prove the absence of republished rows where no public log link exists. The count of rows without public logs is shown on each result.
  • FrontierCode is listed in Epoch's data as an external benchmark whose scores come from Cognition (cognition.com/frontiercode). Epoch's CC BY statement covers Epoch's compilation. We have not verified Cognition's own terms for the underlying figures.
  • FrontierCode rows use different coding harnesses (for example claude-code, codex, devin), so scores are not directly comparable between models. We show them as supporting evidence only and never rate them.

LMArena (Arena) leaderboard dataset

Publisher
LMArena / Arena
Licence
CC-BY-4.0 (terms). Declared on the dataset card (licence: cc-by-4.0). Checked 2026-10-28.
Attribution
Data: LMArena (Arena) leaderboard dataset, lmarena-ai/leaderboard-dataset on Hugging Face, licensed CC BY 4.0.
Benchmarks extracted
Arena text: overall preference, Arena text: creative writing, Arena text: coding, Arena text: maths, Arena text: following instructions, Arena text: business and management, Arena WebDev: building web apps
Last data check
29 Oct 2026, 14:36 UTC
Last verified source update
Newest leaderboard publish date among used results: 28 Oct 2026
Filtering and exclusions
Only the 'latest' leaderboard snapshot of the 'text' and 'webdev' subsets. Rows matched by exact model name and category to config/models.json.
Published here
Yes

Files used

Files used from LMArena (Arena) leaderboard dataset
FileLocationSHA-256 (start)Content first seen
texthttps://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/tree/main/text/latest-00000-of-00001.parquet2f2b2d0891188311…29 Oct 2026
webdevhttps://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/tree/main/webdev/latest-00000-of-00001.parquete2a18304ba61c943…29 Oct 2026

Unresolved questions

  • The dataset card declares one licence for the whole dataset. It does not say whether that resolves the rights to every underlying prompt, vote or model-provider term, and we have not independently confirmed that.
  • We use the 'text' subset (ratings without style control), because that is the subset the project brief approved. Arena made style-controlled ratings its default for text in May 2025, so the numbers here can differ from what arena.ai displays by default.

SWE-bench leaderboard (mini-SWE-agent runs)

Publisher
SWE-bench team
Licence
CC-BY-NC-4.0 (terms). The repository that hosts this file carries a Creative Commons Attribution-NonCommercial 4.0 LICENSE. Checked 2026-10-28.
Attribution
SWE-bench leaderboard data, SWE-bench team (CC BY-NC 4.0).
Benchmarks extracted
None are published (audit copy only).
Last data check
29 Oct 2026, 14:36 UTC
Last verified source update
Not applicable: not published.
Filtering and exclusions
Verified leaderboard, mini-SWE-agent entries only. Excluded from every published number.
Published here
No. The hosting repository's licence is NonCommercial and we cannot establish that this site's use is non-commercial in every future form. In addition, the mini-SWE-agent entries span agent versions 0.0.0 to 2.0.0 with different step and cost budgets, so they are not one uniform test. Affected records are excluded from every published number. They are kept in data/processed/swebench_audit.json for review.

Files used

Files used from SWE-bench leaderboard (mini-SWE-agent runs)
FileLocationSHA-256 (start)Content first seen
leaderboardshttps://raw.githubusercontent.com/SWE-bench/swe-bench.github.io/master/data/leaderboards.json83cd949a9582f4dd…28 Oct 2026

Unresolved questions

  • Owner decision needed: confirm whether this site counts as non-commercial use before enabling publication.
  • If enabled, restrict to a single mini-SWE-agent version so settings, tools and budgets are comparable.

Open issues and exclusions

  • Provenance of Epoch rows. 101 of the 191 Epoch results we publish have no public run log in the data file. The file has no explicit provenance field for these benchmarks, so absence of republished rows is established by marker checks, not proven. Rows excluded by the rule: 0.
  • SWE-bench. 47 mini-SWE-agent runs found and excluded. See the SWE-bench entry above.
  • Models awaiting mapping review. 1 held back. In addition, 8 Arena text, 15 Arena WebDev and 30 Epoch identifiers from recent leaderboards are not mapped to any listed model. They are reported in data/processed/unmapped.json and are not published.
  • Prices. No approved source for consumer subscription prices. Every price reads “Price not independently verified”.
  • Not measured. The tests do not cover tone, brand voice, long documents, spreadsheets, images, speed, privacy terms or reliability over time.

Provenance of every published result

Rating bands, where shown, compare only the 77 listed models. The same table is available as a CSV.

Show all 748 published results and where each came from
Provenance of every published result
ModelBenchmarkResultOriginal identifierConfigurationUnitLicenceSource
amazon-nova-experimental-chat (26-02-10)Arena text: business and management1,450amazon-nova-experimental-chat (26-02-10)26-02-10Arena scoreCC-BY-4.0lmarena
amazon-nova-experimental-chat (26-02-10)Arena text: coding1,483amazon-nova-experimental-chat (26-02-10)26-02-10Arena scoreCC-BY-4.0lmarena
amazon-nova-experimental-chat (26-02-10)Arena text: creative writing1,369amazon-nova-experimental-chat (26-02-10)26-02-10Arena scoreCC-BY-4.0lmarena
amazon-nova-experimental-chat (26-02-10)Arena text: following instructions1,427amazon-nova-experimental-chat (26-02-10)26-02-10Arena scoreCC-BY-4.0lmarena
amazon-nova-experimental-chat (26-02-10)Arena text: maths1,437amazon-nova-experimental-chat (26-02-10)26-02-10Arena scoreCC-BY-4.0lmarena
amazon-nova-experimental-chat (26-02-10)Arena text: overall preference1,449amazon-nova-experimental-chat (26-02-10)26-02-10Arena scoreCC-BY-4.0lmarena
Claude Fable 5Arena text: business and management1,480Claude Fable 5 (High)HighArena scoreCC-BY-4.0lmarena
Claude Fable 5Arena text: coding1,518Claude Fable 5 (High)HighArena scoreCC-BY-4.0lmarena
Claude Fable 5Arena text: creative writing1,493Claude Fable 5 (High)HighArena scoreCC-BY-4.0lmarena
Claude Fable 5Arena text: following instructions1,503Claude Fable 5 (High)HighArena scoreCC-BY-4.0lmarena
Claude Fable 5Arena text: maths1,516Claude Fable 5 (High)HighArena scoreCC-BY-4.0lmarena
Claude Fable 5Arena text: overall preference1,492Claude Fable 5 (High)HighArena scoreCC-BY-4.0lmarena
Claude Fable 5Arena WebDev: building web apps1,624Claude Fable 5 (High)HighArena scoreCC-BY-4.0lmarena
Claude Fable 5OTIS Mock AIME 2024-202599.7%claude-fable-5_maxmaxshare of problems solvedCC-BY-4.0epoch
Claude Fable 5FrontierCode (reported by Cognition)53.5%claude-fable-5_unknownxhigh, claude-codemean score over 5 attemptsCC-BY-4.0epoch
Claude Fable 5FrontierMath Tiers 1-3 (v2)87.0%claude-fable-5_maxmaxshare of problems solvedCC-BY-4.0epoch
Claude Fable 5GPQA Diamond85.9%claude-fable-5_maxmaxshare of questions correctCC-BY-4.0epoch
Claude Fable 5SimpleQA Verified70.7%claude-fable-5_xhighxhighshare of questions correctCC-BY-4.0epoch
Claude Fable 5.1Arena text: business and management1,494Claude Fable 5.1 (Max)MaxArena scoreCC-BY-4.0lmarena
Claude Fable 5.1Arena text: coding1,530Claude Fable 5.1 (Max)MaxArena scoreCC-BY-4.0lmarena
Claude Fable 5.1Arena text: creative writing1,512Claude Fable 5.1 (Max)MaxArena scoreCC-BY-4.0lmarena
Claude Fable 5.1Arena text: following instructions1,519Claude Fable 5.1 (Max)MaxArena scoreCC-BY-4.0lmarena
Claude Fable 5.1Arena text: maths1,531Claude Fable 5.1 (Max)MaxArena scoreCC-BY-4.0lmarena
Claude Fable 5.1Arena text: overall preference1,514Claude Fable 5.1 (Max)MaxArena scoreCC-BY-4.0lmarena
Claude Fable 5.1Arena WebDev: building web apps1,746Claude Fable 5.1 (Max)MaxArena scoreCC-BY-4.0lmarena
Claude Fable 5.1OTIS Mock AIME 2024-2025100.0%claude-fable-5-1_maxmaxshare of problems solvedCC-BY-4.0epoch
Claude Fable 5.1FrontierCode (reported by Cognition)50.9%claude-fable-5-1_mediummedium, claude-codemean score over 5 attemptsCC-BY-4.0epoch
Claude Fable 5.1FrontierMath Tiers 1-3 (v2)90.2%claude-fable-5-1_maxmaxshare of problems solvedCC-BY-4.0epoch
Claude Fable 5.1SimpleQA Verified70.8%claude-fable-5-1_maxmaxshare of questions correctCC-BY-4.0epoch
Claude Haiku 5.5Arena text: business and management1,405Claude Haiku 5.5 (High)HighArena scoreCC-BY-4.0lmarena
Claude Haiku 5.5Arena text: coding1,454Claude Haiku 5.5 (High)HighArena scoreCC-BY-4.0lmarena
Claude Haiku 5.5Arena text: creative writing1,336Claude Haiku 5.5 (High)HighArena scoreCC-BY-4.0lmarena
Claude Haiku 5.5Arena text: following instructions1,415Claude Haiku 5.5 (High)HighArena scoreCC-BY-4.0lmarena
Claude Haiku 5.5Arena text: maths1,434Claude Haiku 5.5 (High)HighArena scoreCC-BY-4.0lmarena
Claude Haiku 5.5Arena text: overall preference1,401Claude Haiku 5.5 (High)HighArena scoreCC-BY-4.0lmarena
Claude Haiku 5.5Arena WebDev: building web apps1,589Claude Haiku 5.5 (High)HighArena scoreCC-BY-4.0lmarena
Claude Haiku 5.5OTIS Mock AIME 2024-202598.9%claude-haiku-5-5_xhighxhighshare of problems solvedCC-BY-4.0epoch
Claude Haiku 5.5FrontierCode (reported by Cognition)46.4%claude-haiku-5-5_maxmax, claude-codemean score over 5 attemptsCC-BY-4.0epoch
Claude Haiku 5.5FrontierMath Tiers 1-3 (v2)75.1%claude-haiku-5-5_maxmaxshare of problems solvedCC-BY-4.0epoch
Claude Haiku 5.5GPQA Diamond89.6%claude-haiku-5-5_maxmaxshare of questions correctCC-BY-4.0epoch
Claude Haiku 5.5SimpleQA Verified23.8%claude-haiku-5-5_maxmaxshare of questions correctCC-BY-4.0epoch
Claude Opus 4.5 (20251101)Arena text: business and management1,448Claude Opus 4.5 (20251101)20251101Arena scoreCC-BY-4.0lmarena
Claude Opus 4.5 (20251101)Arena text: coding1,498Claude Opus 4.5 (20251101)20251101Arena scoreCC-BY-4.0lmarena
Claude Opus 4.5 (20251101)Arena text: creative writing1,443Claude Opus 4.5 (20251101)20251101Arena scoreCC-BY-4.0lmarena
Claude Opus 4.5 (20251101)Arena text: following instructions1,472Claude Opus 4.5 (20251101)20251101Arena scoreCC-BY-4.0lmarena
Claude Opus 4.5 (20251101)Arena text: maths1,458Claude Opus 4.5 (20251101)20251101Arena scoreCC-BY-4.0lmarena
Claude Opus 4.5 (20251101)Arena text: overall preference1,451Claude Opus 4.5 (20251101)20251101Arena scoreCC-BY-4.0lmarena
Claude Opus 4.5 (20251101)Arena WebDev: building web apps1,469Claude Opus 4.5 (20251101)20251101Arena scoreCC-BY-4.0lmarena
Claude Opus 4.6Arena text: business and management1,497Claude Opus 4.6 (High)HighArena scoreCC-BY-4.0lmarena
Claude Opus 4.6Arena text: coding1,531Claude Opus 4.6 (High)HighArena scoreCC-BY-4.0lmarena
Claude Opus 4.6Arena text: creative writing1,504Claude Opus 4.6 (High)HighArena scoreCC-BY-4.0lmarena
Claude Opus 4.6Arena text: following instructions1,521Claude Opus 4.6 (High)HighArena scoreCC-BY-4.0lmarena
Claude Opus 4.6Arena text: maths1,518Claude Opus 4.6 (High)HighArena scoreCC-BY-4.0lmarena
Claude Opus 4.6Arena text: overall preference1,502Claude Opus 4.6 (High)HighArena scoreCC-BY-4.0lmarena
Claude Opus 4.6Arena WebDev: building web apps1,546Claude Opus 4.6 (High)HighArena scoreCC-BY-4.0lmarena
Claude Opus 4.6OTIS Mock AIME 2024-202591.1%claude-opus-4-6_maxmaxshare of problems solvedCC-BY-4.0epoch
Claude Opus 4.6FrontierCode (reported by Cognition)26.6%claude-opus-4-6_unknownhigh, claude-codemean score over 5 attemptsCC-BY-4.0epoch
Claude Opus 4.6FrontierMath Tiers 1-3 (v2)66.0%claude-opus-4-6_maxmaxshare of problems solvedCC-BY-4.0epoch
Claude Opus 4.6GPQA Diamond88.4%claude-opus-4-6_maxmaxshare of questions correctCC-BY-4.0epoch
Claude Opus 4.6SimpleQA Verified47.0%claude-opus-4-6_maxmaxshare of questions correctCC-BY-4.0epoch
Claude Opus 4.7Arena text: business and management1,490Claude Opus 4.7 (High)HighArena scoreCC-BY-4.0lmarena
Claude Opus 4.7Arena text: coding1,519Claude Opus 4.7 (High)HighArena scoreCC-BY-4.0lmarena
Claude Opus 4.7Arena text: creative writing1,486Claude Opus 4.7 (High)HighArena scoreCC-BY-4.0lmarena
Claude Opus 4.7Arena text: following instructions1,498Claude Opus 4.7 (High)HighArena scoreCC-BY-4.0lmarena
Claude Opus 4.7Arena text: maths1,499Claude Opus 4.7 (High)HighArena scoreCC-BY-4.0lmarena
Claude Opus 4.7Arena text: overall preference1,490Claude Opus 4.7 (High)HighArena scoreCC-BY-4.0lmarena
Claude Opus 4.7Arena WebDev: building web apps1,558Claude Opus 4.7 (High)HighArena scoreCC-BY-4.0lmarena
Claude Opus 4.7OTIS Mock AIME 2024-202586.7%claude-opus-4-7_maxmaxshare of problems solvedCC-BY-4.0epoch
Claude Opus 4.7FrontierCode (reported by Cognition)38.5%claude-opus-4-7_unknownmax, claude-codemean score over 5 attemptsCC-BY-4.0epoch
Claude Opus 4.7FrontierMath Tiers 1-3 (v2)70.2%claude-opus-4-7_maxmaxshare of problems solvedCC-BY-4.0epoch
Claude Opus 4.7GPQA Diamond86.4%claude-opus-4-7_maxmaxshare of questions correctCC-BY-4.0epoch
Claude Opus 4.7SimpleQA Verified51.7%claude-opus-4-7_xhighxhighshare of questions correctCC-BY-4.0epoch
Claude Opus 4.7SWE-bench Verified (Epoch runs)83.5%claude-opus-4-7_maxmaxshare of issues resolvedCC-BY-4.0epoch
Claude Opus 4.8Arena text: business and management1,458Claude Opus 4.8 (High)HighArena scoreCC-BY-4.0lmarena
Claude Opus 4.8Arena text: coding1,491Claude Opus 4.8 (High)HighArena scoreCC-BY-4.0lmarena
Claude Opus 4.8Arena text: creative writing1,455Claude Opus 4.8 (High)HighArena scoreCC-BY-4.0lmarena
Claude Opus 4.8Arena text: following instructions1,477Claude Opus 4.8 (High)HighArena scoreCC-BY-4.0lmarena
Claude Opus 4.8Arena text: maths1,485Claude Opus 4.8 (High)HighArena scoreCC-BY-4.0lmarena
Claude Opus 4.8Arena text: overall preference1,461Claude Opus 4.8 (High)HighArena scoreCC-BY-4.0lmarena
Claude Opus 4.8Arena WebDev: building web apps1,556Claude Opus 4.8 (High)HighArena scoreCC-BY-4.0lmarena
Claude Opus 4.8OTIS Mock AIME 2024-202598.3%claude-opus-4-8_maxmaxshare of problems solvedCC-BY-4.0epoch
Claude Opus 4.8FrontierCode (reported by Cognition)46.5%claude-opus-4-8_unknownmax, claude-codemean score over 5 attemptsCC-BY-4.0epoch
Claude Opus 4.8FrontierMath Tiers 1-3 (v2)80.0%claude-opus-4-8_maxmaxshare of problems solvedCC-BY-4.0epoch
Claude Opus 4.8GPQA Diamond91.0%claude-opus-4-8_maxmaxshare of questions correctCC-BY-4.0epoch
Claude Opus 4.8SimpleQA Verified53.0%claude-opus-4-8_maxmaxshare of questions correctCC-BY-4.0epoch
Claude Opus 5Arena text: business and management1,500Claude Opus 5 (Max)MaxArena scoreCC-BY-4.0lmarena
Claude Opus 5Arena text: coding1,532Claude Opus 5 (Max)MaxArena scoreCC-BY-4.0lmarena
Claude Opus 5Arena text: creative writing1,495Claude Opus 5 (Max)MaxArena scoreCC-BY-4.0lmarena
Claude Opus 5Arena text: following instructions1,518Claude Opus 5 (Max)MaxArena scoreCC-BY-4.0lmarena
Claude Opus 5Arena text: maths1,531Claude Opus 5 (Max)MaxArena scoreCC-BY-4.0lmarena
Claude Opus 5Arena text: overall preference1,508Claude Opus 5 (Max)MaxArena scoreCC-BY-4.0lmarena
Claude Opus 5Arena WebDev: building web apps1,691Claude Opus 5 (Max)MaxArena scoreCC-BY-4.0lmarena
Claude Opus 5OTIS Mock AIME 2024-202598.9%claude-opus-5_maxmaxshare of problems solvedCC-BY-4.0epoch
Claude Opus 5FrontierCode (reported by Cognition)53.4%claude-opus-5_maxmedium, claude-codemean score over 5 attemptsCC-BY-4.0epoch
Claude Opus 5FrontierMath Tiers 1-3 (v2)85.6%claude-opus-5_maxmaxshare of problems solvedCC-BY-4.0epoch
Claude Opus 5GPQA Diamond93.9%claude-opus-5_maxmaxshare of questions correctCC-BY-4.0epoch
Claude Opus 5SimpleQA Verified59.9%claude-opus-5_maxmaxshare of questions correctCC-BY-4.0epoch
Claude Opus 5.5Arena text: business and management1,497Claude Opus 5.5 (High)HighArena scoreCC-BY-4.0lmarena
Claude Opus 5.5Arena text: coding1,547Claude Opus 5.5 (High)HighArena scoreCC-BY-4.0lmarena
Claude Opus 5.5Arena text: creative writing1,534Claude Opus 5.5 (High)HighArena scoreCC-BY-4.0lmarena
Claude Opus 5.5Arena text: following instructions1,539Claude Opus 5.5 (High)HighArena scoreCC-BY-4.0lmarena
Claude Opus 5.5Arena text: maths1,517Claude Opus 5.5 (High)HighArena scoreCC-BY-4.0lmarena
Claude Opus 5.5Arena text: overall preference1,516Claude Opus 5.5 (High)HighArena scoreCC-BY-4.0lmarena
Claude Opus 5.5Arena WebDev: building web apps1,810Claude Opus 5.5 (Max)MaxArena scoreCC-BY-4.0lmarena
Claude Opus 5.5OTIS Mock AIME 2024-2025100.0%claude-opus-5-5_maxmaxshare of problems solvedCC-BY-4.0epoch
Claude Opus 5.5FrontierCode (reported by Cognition)54.6%claude-opus-5-5_mediummedium, claude-codemean score over 5 attemptsCC-BY-4.0epoch
Claude Opus 5.5FrontierMath Tiers 1-3 (v2)91.2%claude-opus-5-5_maxmaxshare of problems solvedCC-BY-4.0epoch
Claude Opus 5.5GPQA Diamond90.6%claude-opus-5-5_maxmaxshare of questions correctCC-BY-4.0epoch
Claude Opus 5.5SimpleQA Verified72.2%claude-opus-5-5_maxmaxshare of questions correctCC-BY-4.0epoch
Claude Sonnet 4.6Arena text: business and management1,464Claude Sonnet 4.6not statedArena scoreCC-BY-4.0lmarena
Claude Sonnet 4.6Arena text: coding1,503Claude Sonnet 4.6not statedArena scoreCC-BY-4.0lmarena
Claude Sonnet 4.6Arena text: creative writing1,435Claude Sonnet 4.6not statedArena scoreCC-BY-4.0lmarena
Claude Sonnet 4.6Arena text: following instructions1,476Claude Sonnet 4.6not statedArena scoreCC-BY-4.0lmarena
Claude Sonnet 4.6Arena text: maths1,462Claude Sonnet 4.6not statedArena scoreCC-BY-4.0lmarena
Claude Sonnet 4.6Arena text: overall preference1,458Claude Sonnet 4.6not statedArena scoreCC-BY-4.0lmarena
Claude Sonnet 4.6Arena WebDev: building web apps1,523Claude Sonnet 4.6not statedArena scoreCC-BY-4.0lmarena
Claude Sonnet 4.6OTIS Mock AIME 2024-202571.1%claude-sonnet-4-6_maxmaxshare of problems solvedCC-BY-4.0epoch
Claude Sonnet 4.6FrontierCode (reported by Cognition)24.3%claude-sonnet-4-6_unknownmax, claude-codemean score over 5 attemptsCC-BY-4.0epoch
Claude Sonnet 4.6GPQA Diamond78.8%claude-sonnet-4-6_maxmaxshare of questions correctCC-BY-4.0epoch
Claude Sonnet 4.6SimpleQA Verified32.8%claude-sonnet-4-6_maxmaxshare of questions correctCC-BY-4.0epoch
Claude Sonnet 4.6SWE-bench Verified (Epoch runs)75.2%claude-sonnet-4-6not statedshare of issues resolvedCC-BY-4.0epoch
Claude Sonnet 5Arena text: business and management1,437Claude Sonnet 5 (High)HighArena scoreCC-BY-4.0lmarena
Claude Sonnet 5Arena text: coding1,484Claude Sonnet 5 (High)HighArena scoreCC-BY-4.0lmarena
Claude Sonnet 5Arena text: creative writing1,416Claude Sonnet 5 (High)HighArena scoreCC-BY-4.0lmarena
Claude Sonnet 5Arena text: following instructions1,453Claude Sonnet 5 (High)HighArena scoreCC-BY-4.0lmarena
Claude Sonnet 5Arena text: maths1,469Claude Sonnet 5 (High)HighArena scoreCC-BY-4.0lmarena
Claude Sonnet 5Arena text: overall preference1,443Claude Sonnet 5 (High)HighArena scoreCC-BY-4.0lmarena
Claude Sonnet 5Arena WebDev: building web apps1,540Claude Sonnet 5 (High)HighArena scoreCC-BY-4.0lmarena
Claude Sonnet 5OTIS Mock AIME 2024-202580.0%claude-sonnet-5_maxmaxshare of problems solvedCC-BY-4.0epoch
Claude Sonnet 5FrontierCode (reported by Cognition)42.7%claude-sonnet-5_unknownxhigh, claude-codemean score over 5 attemptsCC-BY-4.0epoch
Claude Sonnet 5FrontierMath Tiers 1-3 (v2)65.6%claude-sonnet-5_maxmaxshare of problems solvedCC-BY-4.0epoch
Claude Sonnet 5GPQA Diamond80.3%claude-sonnet-5_maxmaxshare of questions correctCC-BY-4.0epoch
Claude Sonnet 5SimpleQA Verified33.7%claude-sonnet-5_maxmaxshare of questions correctCC-BY-4.0epoch
Claude Sonnet 5.5Arena text: business and management1,466Claude Sonnet 5.5 (xHigh)xHighArena scoreCC-BY-4.0lmarena
Claude Sonnet 5.5Arena text: coding1,512Claude Sonnet 5.5 (xHigh)xHighArena scoreCC-BY-4.0lmarena
Claude Sonnet 5.5Arena text: creative writing1,470Claude Sonnet 5.5 (xHigh)xHighArena scoreCC-BY-4.0lmarena
Claude Sonnet 5.5Arena text: following instructions1,492Claude Sonnet 5.5 (xHigh)xHighArena scoreCC-BY-4.0lmarena
Claude Sonnet 5.5Arena text: maths1,504Claude Sonnet 5.5 (xHigh)xHighArena scoreCC-BY-4.0lmarena
Claude Sonnet 5.5Arena text: overall preference1,473Claude Sonnet 5.5 (xHigh)xHighArena scoreCC-BY-4.0lmarena
Claude Sonnet 5.5Arena WebDev: building web apps1,770Claude Sonnet 5.5 (xHigh)xHighArena scoreCC-BY-4.0lmarena
Claude Sonnet 5.5OTIS Mock AIME 2024-2025100.0%claude-sonnet-5-5_maxmaxshare of problems solvedCC-BY-4.0epoch
Claude Sonnet 5.5FrontierCode (reported by Cognition)52.1%claude-sonnet-5-5_xhighxhigh, claude-codemean score over 5 attemptsCC-BY-4.0epoch
Claude Sonnet 5.5FrontierMath Tiers 1-3 (v2)88.8%claude-sonnet-5-5_maxmaxshare of problems solvedCC-BY-4.0epoch
Claude Sonnet 5.5GPQA Diamond95.6%claude-sonnet-5-5_maxmaxshare of questions correctCC-BY-4.0epoch
Claude Sonnet 5.5SimpleQA Verified46.5%claude-sonnet-5-5_maxmaxshare of questions correctCC-BY-4.0epoch
DeepSeek-V4.1-Flash-20260910Arena text: business and management1,462DeepSeek-V4.1-Flash-20260910 (Max)MaxArena scoreCC-BY-4.0lmarena
DeepSeek-V4.1-Flash-20260910Arena text: coding1,501DeepSeek-V4.1-Flash-20260910 (Max)MaxArena scoreCC-BY-4.0lmarena
DeepSeek-V4.1-Flash-20260910Arena text: creative writing1,437DeepSeek-V4.1-Flash-20260910 (Max)MaxArena scoreCC-BY-4.0lmarena
DeepSeek-V4.1-Flash-20260910Arena text: following instructions1,474DeepSeek-V4.1-Flash-20260910 (Max)MaxArena scoreCC-BY-4.0lmarena
DeepSeek-V4.1-Flash-20260910Arena text: maths1,481DeepSeek-V4.1-Flash-20260910 (Max)MaxArena scoreCC-BY-4.0lmarena
DeepSeek-V4.1-Flash-20260910Arena text: overall preference1,465DeepSeek-V4.1-Flash-20260910 (Max)MaxArena scoreCC-BY-4.0lmarena
DeepSeek-V4.1-Flash-20260910Arena WebDev: building web apps1,619DeepSeek-V4.1-Flash-20260910 (Max)MaxArena scoreCC-BY-4.0lmarena
DeepSeek-V4-ProArena text: business and management1,443DeepSeek-V4-Pronot statedArena scoreCC-BY-4.0lmarena
DeepSeek-V4-ProArena text: coding1,469DeepSeek-V4-Pronot statedArena scoreCC-BY-4.0lmarena
DeepSeek-V4-ProArena text: creative writing1,446DeepSeek-V4-Pronot statedArena scoreCC-BY-4.0lmarena
DeepSeek-V4-ProArena text: following instructions1,445DeepSeek-V4-Pronot statedArena scoreCC-BY-4.0lmarena
DeepSeek-V4-ProArena text: maths1,442DeepSeek-V4-Pronot statedArena scoreCC-BY-4.0lmarena
DeepSeek-V4-ProArena text: overall preference1,451DeepSeek-V4-Pronot statedArena scoreCC-BY-4.0lmarena
DeepSeek-V4-ProArena WebDev: building web apps1,447DeepSeek-V4-Pronot statedArena scoreCC-BY-4.0lmarena
DeepSeek-V4-ProOTIS Mock AIME 2024-202596.7%deepseek-v4-pro_maxmaxshare of problems solvedCC-BY-4.0epoch
DeepSeek-V4-ProFrontierCode (reported by Cognition)17.6%deepseek-v4-pro_nonenone, mini-swe-agentmean score over 5 attemptsCC-BY-4.0epoch
DeepSeek-V4-ProFrontierMath Tiers 1-3 (v2)45.3%deepseek-v4-pro_maxmaxshare of problems solvedCC-BY-4.0epoch
DeepSeek-V4-ProGPQA Diamond89.6%deepseek-v4-pro_maxmaxshare of questions correctCC-BY-4.0epoch
DeepSeek-V4-ProSimpleQA Verified47.0%deepseek-v4-pro_maxmaxshare of questions correctCC-BY-4.0epoch
DeepSeek-V4-ProSWE-bench Verified (Epoch runs)77.6%deepseek-v4-pro_maxmaxshare of issues resolvedCC-BY-4.0epoch
DeepSeek-V4-Pro (High, 20260813)Arena text: business and management1,442DeepSeek-V4-Pro (High, 20260813)High, 20260813Arena scoreCC-BY-4.0lmarena
DeepSeek-V4-Pro (High, 20260813)Arena text: coding1,469DeepSeek-V4-Pro (High, 20260813)High, 20260813Arena scoreCC-BY-4.0lmarena
DeepSeek-V4-Pro (High, 20260813)Arena text: creative writing1,440DeepSeek-V4-Pro (High, 20260813)High, 20260813Arena scoreCC-BY-4.0lmarena
DeepSeek-V4-Pro (High, 20260813)Arena text: following instructions1,449DeepSeek-V4-Pro (High, 20260813)High, 20260813Arena scoreCC-BY-4.0lmarena
DeepSeek-V4-Pro (High, 20260813)Arena text: maths1,442DeepSeek-V4-Pro (High, 20260813)High, 20260813Arena scoreCC-BY-4.0lmarena
DeepSeek-V4-Pro (High, 20260813)Arena text: overall preference1,446DeepSeek-V4-Pro (High, 20260813)High, 20260813Arena scoreCC-BY-4.0lmarena
DeepSeek-V4-Pro (High, 20260813)Arena WebDev: building web apps1,582DeepSeek-V4-Pro (High, 20260813)High, 20260813Arena scoreCC-BY-4.0lmarena
DeepSeek-V4-Pro (High, 20260813)OTIS Mock AIME 2024-202598.6%deepseek-v4-pro-0813_maxmaxshare of problems solvedCC-BY-4.0epoch
DeepSeek-V4-Pro (High, 20260813)FrontierCode (reported by Cognition)28.5%deepseek-v4-pro-0813_highhigh, devin-climean score over 5 attemptsCC-BY-4.0epoch
DeepSeek-V4-Pro (High, 20260813)FrontierMath Tiers 1-3 (v2)64.6%deepseek-v4-pro-0813_maxmaxshare of problems solvedCC-BY-4.0epoch
DeepSeek-V4-Pro (High, 20260813)GPQA Diamond91.7%deepseek-v4-pro-0813_maxmaxshare of questions correctCC-BY-4.0epoch
DeepSeek-V4-Pro (High, 20260813)SimpleQA Verified52.9%deepseek-v4-pro-0813_maxmaxshare of questions correctCC-BY-4.0epoch
DeepSeek-V4-Pro PreviewArena text: business and management1,428DeepSeek-V4-Pro Preview (High)HighArena scoreCC-BY-4.0lmarena
DeepSeek-V4-Pro PreviewArena text: coding1,453DeepSeek-V4-Pro Preview (High)HighArena scoreCC-BY-4.0lmarena
DeepSeek-V4-Pro PreviewArena text: creative writing1,441DeepSeek-V4-Pro Preview (High)HighArena scoreCC-BY-4.0lmarena
DeepSeek-V4-Pro PreviewArena text: following instructions1,436DeepSeek-V4-Pro Preview (High)HighArena scoreCC-BY-4.0lmarena
DeepSeek-V4-Pro PreviewArena text: maths1,454DeepSeek-V4-Pro Preview (High)HighArena scoreCC-BY-4.0lmarena
DeepSeek-V4-Pro PreviewArena text: overall preference1,445DeepSeek-V4-Pro Preview (High)HighArena scoreCC-BY-4.0lmarena
DeepSeek-V4-Pro PreviewArena WebDev: building web apps1,464DeepSeek-V4-Pro Preview (High)HighArena scoreCC-BY-4.0lmarena
Dola Seed 2.0 ProArena text: business and management1,439Dola Seed 2.0 Pronot statedArena scoreCC-BY-4.0lmarena
Dola Seed 2.0 ProArena text: coding1,472Dola Seed 2.0 Pronot statedArena scoreCC-BY-4.0lmarena
Dola Seed 2.0 ProArena text: creative writing1,406Dola Seed 2.0 Pronot statedArena scoreCC-BY-4.0lmarena
Dola Seed 2.0 ProArena text: following instructions1,414Dola Seed 2.0 Pronot statedArena scoreCC-BY-4.0lmarena
Dola Seed 2.0 ProArena text: maths1,439Dola Seed 2.0 Pronot statedArena scoreCC-BY-4.0lmarena
Dola Seed 2.0 ProArena text: overall preference1,448Dola Seed 2.0 Pronot statedArena scoreCC-BY-4.0lmarena
ERNIE 5.0 (0110)Arena text: business and management1,429ERNIE 5.0 (0110)0110Arena scoreCC-BY-4.0lmarena
ERNIE 5.0 (0110)Arena text: coding1,455ERNIE 5.0 (0110)0110Arena scoreCC-BY-4.0lmarena
ERNIE 5.0 (0110)Arena text: creative writing1,426ERNIE 5.0 (0110)0110Arena scoreCC-BY-4.0lmarena
ERNIE 5.0 (0110)Arena text: following instructions1,413ERNIE 5.0 (0110)0110Arena scoreCC-BY-4.0lmarena
ERNIE 5.0 (0110)Arena text: maths1,436ERNIE 5.0 (0110)0110Arena scoreCC-BY-4.0lmarena
ERNIE 5.0 (0110)Arena text: overall preference1,445ERNIE 5.0 (0110)0110Arena scoreCC-BY-4.0lmarena
ERNIE 5.1Arena text: business and management1,473ERNIE 5.1not statedArena scoreCC-BY-4.0lmarena
ERNIE 5.1Arena text: coding1,488ERNIE 5.1not statedArena scoreCC-BY-4.0lmarena
ERNIE 5.1Arena text: creative writing1,441ERNIE 5.1not statedArena scoreCC-BY-4.0lmarena
ERNIE 5.1Arena text: following instructions1,460ERNIE 5.1not statedArena scoreCC-BY-4.0lmarena
ERNIE 5.1Arena text: maths1,482ERNIE 5.1not statedArena scoreCC-BY-4.0lmarena
ERNIE 5.1Arena text: overall preference1,468ERNIE 5.1not statedArena scoreCC-BY-4.0lmarena
Gemini 2.5 ProArena text: business and management1,439Gemini 2.5 Pronot statedArena scoreCC-BY-4.0lmarena
Gemini 2.5 ProArena text: coding1,451Gemini 2.5 Pronot statedArena scoreCC-BY-4.0lmarena
Gemini 2.5 ProArena text: creative writing1,454Gemini 2.5 Pronot statedArena scoreCC-BY-4.0lmarena
Gemini 2.5 ProArena text: following instructions1,437Gemini 2.5 Pronot statedArena scoreCC-BY-4.0lmarena
Gemini 2.5 ProArena text: maths1,450Gemini 2.5 Pronot statedArena scoreCC-BY-4.0lmarena
Gemini 2.5 ProArena text: overall preference1,458Gemini 2.5 Pronot statedArena scoreCC-BY-4.0lmarena
Gemini 2.5 ProArena WebDev: building web apps1,226Gemini 2.5 Pronot statedArena scoreCC-BY-4.0lmarena
Gemini 3.1 Pro PreviewArena text: business and management1,459Gemini 3.1 Pro Previewnot statedArena scoreCC-BY-4.0lmarena
Gemini 3.1 Pro PreviewArena text: coding1,484Gemini 3.1 Pro Previewnot statedArena scoreCC-BY-4.0lmarena
Gemini 3.1 Pro PreviewArena text: creative writing1,482Gemini 3.1 Pro Previewnot statedArena scoreCC-BY-4.0lmarena
Gemini 3.1 Pro PreviewArena text: following instructions1,466Gemini 3.1 Pro Previewnot statedArena scoreCC-BY-4.0lmarena
Gemini 3.1 Pro PreviewArena text: maths1,485Gemini 3.1 Pro Previewnot statedArena scoreCC-BY-4.0lmarena
Gemini 3.1 Pro PreviewArena text: overall preference1,480Gemini 3.1 Pro Previewnot statedArena scoreCC-BY-4.0lmarena
Gemini 3.1 Pro PreviewArena WebDev: building web apps1,447Gemini 3.1 Pro Previewnot statedArena scoreCC-BY-4.0lmarena
Gemini 3.5 FlashArena text: business and management1,467Gemini 3.5 Flash (High)HighArena scoreCC-BY-4.0lmarena
Gemini 3.5 FlashArena text: coding1,493Gemini 3.5 Flash (High)HighArena scoreCC-BY-4.0lmarena
Gemini 3.5 FlashArena text: creative writing1,469Gemini 3.5 Flash (High)HighArena scoreCC-BY-4.0lmarena
Gemini 3.5 FlashArena text: following instructions1,467Gemini 3.5 Flash (High)HighArena scoreCC-BY-4.0lmarena
Gemini 3.5 FlashArena text: maths1,504Gemini 3.5 Flash (High)HighArena scoreCC-BY-4.0lmarena
Gemini 3.5 FlashArena text: overall preference1,482Gemini 3.5 Flash (High)HighArena scoreCC-BY-4.0lmarena
Gemini 3.5 FlashArena WebDev: building web apps1,499Gemini 3.5 Flash (High)HighArena scoreCC-BY-4.0lmarena
Gemini 3.5 FlashOTIS Mock AIME 2024-202595.6%gemini-3.5-flash_highhighshare of problems solvedCC-BY-4.0epoch
Gemini 3.5 FlashFrontierMath Tiers 1-3 (v2)62.8%gemini-3.5-flash_highhighshare of problems solvedCC-BY-4.0epoch
Gemini 3.5 FlashGPQA Diamond92.8%gemini-3.5-flash_highhighshare of questions correctCC-BY-4.0epoch
Gemini 3.5 FlashSimpleQA Verified66.2%gemini-3.5-flash_highhighshare of questions correctCC-BY-4.0epoch
Gemini 3.5 FlashSWE-bench Verified (Epoch runs)79.3%gemini-3.5-flash_highhighshare of issues resolvedCC-BY-4.0epoch
Gemini 3.5 Flash-LiteArena text: business and management1,423Gemini 3.5 Flash-Litenot statedArena scoreCC-BY-4.0lmarena
Gemini 3.5 Flash-LiteArena text: coding1,453Gemini 3.5 Flash-Litenot statedArena scoreCC-BY-4.0lmarena
Gemini 3.5 Flash-LiteArena text: creative writing1,421Gemini 3.5 Flash-Litenot statedArena scoreCC-BY-4.0lmarena
Gemini 3.5 Flash-LiteArena text: following instructions1,420Gemini 3.5 Flash-Litenot statedArena scoreCC-BY-4.0lmarena
Gemini 3.5 Flash-LiteArena text: maths1,436Gemini 3.5 Flash-Litenot statedArena scoreCC-BY-4.0lmarena
Gemini 3.5 Flash-LiteArena text: overall preference1,435Gemini 3.5 Flash-Litenot statedArena scoreCC-BY-4.0lmarena
Gemini 3.5 Flash-LiteArena WebDev: building web apps1,434Gemini 3.5 Flash-Litenot statedArena scoreCC-BY-4.0lmarena
Gemini 3.5 Flash-LiteOTIS Mock AIME 2024-202571.1%gemini-3.5-flash-lite_highhighshare of problems solvedCC-BY-4.0epoch
Gemini 3.5 Flash-LiteFrontierMath Tiers 1-3 (v2)26.0%gemini-3.5-flash-lite_highhighshare of problems solvedCC-BY-4.0epoch
Gemini 3.5 Flash-LiteGPQA Diamond83.3%gemini-3.5-flash-lite_highhighshare of questions correctCC-BY-4.0epoch
Gemini 3.6 FlashArena text: business and management1,464Gemini 3.6 Flash (High)HighArena scoreCC-BY-4.0lmarena
Gemini 3.6 FlashArena text: coding1,491Gemini 3.6 Flash (High)HighArena scoreCC-BY-4.0lmarena
Gemini 3.6 FlashArena text: creative writing1,466Gemini 3.6 Flash (High)HighArena scoreCC-BY-4.0lmarena
Gemini 3.6 FlashArena text: following instructions1,467Gemini 3.6 Flash (High)HighArena scoreCC-BY-4.0lmarena
Gemini 3.6 FlashArena text: maths1,507Gemini 3.6 Flash (High)HighArena scoreCC-BY-4.0lmarena
Gemini 3.6 FlashArena text: overall preference1,480Gemini 3.6 Flash (High)HighArena scoreCC-BY-4.0lmarena
Gemini 3.6 FlashArena WebDev: building web apps1,538Gemini 3.6 Flash (High)HighArena scoreCC-BY-4.0lmarena
Gemini 3.6 FlashOTIS Mock AIME 2024-202594.2%gemini-3.6-flash_highhighshare of problems solvedCC-BY-4.0epoch
Gemini 3.6 FlashFrontierCode (reported by Cognition)34.4%gemini-3.6-flash_unknownmedium, devin-climean score over 5 attemptsCC-BY-4.0epoch
Gemini 3.6 FlashFrontierMath Tiers 1-3 (v2)58.9%gemini-3.6-flash_highhighshare of problems solvedCC-BY-4.0epoch
Gemini 3.6 FlashGPQA Diamond94.1%gemini-3.6-flash_highhighshare of questions correctCC-BY-4.0epoch
Gemini 3.6 FlashSimpleQA Verified66.2%gemini-3.6-flash_highhighshare of questions correctCC-BY-4.0epoch
Gemini 3.7 FlashArena text: business and management1,480Gemini 3.7 Flash (High)HighArena scoreCC-BY-4.0lmarena
Gemini 3.7 FlashArena text: coding1,500Gemini 3.7 Flash (High)HighArena scoreCC-BY-4.0lmarena
Gemini 3.7 FlashArena text: creative writing1,487Gemini 3.7 Flash (High)HighArena scoreCC-BY-4.0lmarena
Gemini 3.7 FlashArena text: following instructions1,486Gemini 3.7 Flash (High)HighArena scoreCC-BY-4.0lmarena
Gemini 3.7 FlashArena text: maths1,507Gemini 3.7 Flash (High)HighArena scoreCC-BY-4.0lmarena
Gemini 3.7 FlashArena text: overall preference1,489Gemini 3.7 Flash (High)HighArena scoreCC-BY-4.0lmarena
Gemini 3.7 FlashArena WebDev: building web apps1,590Gemini 3.7 Flash (High)HighArena scoreCC-BY-4.0lmarena
Gemini 3.7 FlashOTIS Mock AIME 2024-202597.2%gemini-3.7-flash_highhighshare of problems solvedCC-BY-4.0epoch
Gemini 3.7 FlashFrontierCode (reported by Cognition)43.6%gemini-3.7-flash_unknownmedium, devin-climean score over 5 attemptsCC-BY-4.0epoch
Gemini 3.7 FlashFrontierMath Tiers 1-3 (v2)71.6%gemini-3.7-flash_highhighshare of problems solvedCC-BY-4.0epoch
Gemini 3.7 FlashGPQA Diamond94.8%gemini-3.7-flash_highhighshare of questions correctCC-BY-4.0epoch
Gemini 3.7 FlashSimpleQA Verified69.2%gemini-3.7-flash_highhighshare of questions correctCC-BY-4.0epoch
Gemini 3.8 FlashArena text: business and management1,480Gemini 3.8 Flash (High)HighArena scoreCC-BY-4.0lmarena
Gemini 3.8 FlashArena text: coding1,509Gemini 3.8 Flash (High)HighArena scoreCC-BY-4.0lmarena
Gemini 3.8 FlashArena text: creative writing1,491Gemini 3.8 Flash (High)HighArena scoreCC-BY-4.0lmarena
Gemini 3.8 FlashArena text: following instructions1,493Gemini 3.8 Flash (High)HighArena scoreCC-BY-4.0lmarena
Gemini 3.8 FlashArena text: maths1,530Gemini 3.8 Flash (High)HighArena scoreCC-BY-4.0lmarena
Gemini 3.8 FlashArena text: overall preference1,498Gemini 3.8 Flash (High)HighArena scoreCC-BY-4.0lmarena
Gemini 3.8 FlashArena WebDev: building web apps1,583Gemini 3.8 Flash (High)HighArena scoreCC-BY-4.0lmarena
Gemini 3.8 FlashOTIS Mock AIME 2024-202598.9%gemini-3.8-flash_highhighshare of problems solvedCC-BY-4.0epoch
Gemini 3.8 FlashFrontierCode (reported by Cognition)41.2%gemini-3.8-flash_mediummedium, devin-climean score over 5 attemptsCC-BY-4.0epoch
Gemini 3.8 FlashFrontierMath Tiers 1-3 (v2)68.4%gemini-3.8-flash_highhighshare of problems solvedCC-BY-4.0epoch
Gemini 3.8 FlashGPQA Diamond95.4%gemini-3.8-flash_highhighshare of questions correctCC-BY-4.0epoch
Gemini 3.8 FlashSimpleQA Verified69.7%gemini-3.8-flash_highhighshare of questions correctCC-BY-4.0epoch
Gemini 3 FlashArena text: business and management1,447Gemini 3 Flashnot statedArena scoreCC-BY-4.0lmarena
Gemini 3 FlashArena text: coding1,459Gemini 3 Flashnot statedArena scoreCC-BY-4.0lmarena
Gemini 3 FlashArena text: creative writing1,457Gemini 3 Flashnot statedArena scoreCC-BY-4.0lmarena
Gemini 3 FlashArena text: following instructions1,436Gemini 3 Flashnot statedArena scoreCC-BY-4.0lmarena
Gemini 3 FlashArena text: maths1,473Gemini 3 Flashnot statedArena scoreCC-BY-4.0lmarena
Gemini 3 FlashArena text: overall preference1,466Gemini 3 Flashnot statedArena scoreCC-BY-4.0lmarena
Gemini 3 FlashArena WebDev: building web apps1,440Gemini 3 Flashnot statedArena scoreCC-BY-4.0lmarena
Gemini 3 ProArena text: business and management1,458Gemini 3 Pronot statedArena scoreCC-BY-4.0lmarena
Gemini 3 ProArena text: coding1,481Gemini 3 Pronot statedArena scoreCC-BY-4.0lmarena
Gemini 3 ProArena text: creative writing1,482Gemini 3 Pronot statedArena scoreCC-BY-4.0lmarena
Gemini 3 ProArena text: following instructions1,458Gemini 3 Pronot statedArena scoreCC-BY-4.0lmarena
Gemini 3 ProArena text: maths1,476Gemini 3 Pronot statedArena scoreCC-BY-4.0lmarena
Gemini 3 ProArena text: overall preference1,479Gemini 3 Pronot statedArena scoreCC-BY-4.0lmarena
Gemini 3 ProArena WebDev: building web apps1,440Gemini 3 Pronot statedArena scoreCC-BY-4.0lmarena
Gemini 4 ArgonArena text: business and management1,515Gemini 4 Argon (High)HighArena scoreCC-BY-4.0lmarena
Gemini 4 ArgonArena text: coding1,538Gemini 4 Argon (High)HighArena scoreCC-BY-4.0lmarena
Gemini 4 ArgonArena text: creative writing1,527Gemini 4 Argon (High)HighArena scoreCC-BY-4.0lmarena
Gemini 4 ArgonArena text: following instructions1,533Gemini 4 Argon (High)HighArena scoreCC-BY-4.0lmarena
Gemini 4 ArgonArena text: maths1,538Gemini 4 Argon (High)HighArena scoreCC-BY-4.0lmarena
Gemini 4 ArgonArena text: overall preference1,529Gemini 4 Argon (High)HighArena scoreCC-BY-4.0lmarena
Gemini 4 ArgonArena WebDev: building web apps1,679Gemini 4 Argon (High)HighArena scoreCC-BY-4.0lmarena
Gemini 4 ArgonOTIS Mock AIME 2024-2025100.0%gemini-4-argon_highhighshare of problems solvedCC-BY-4.0epoch
Gemini 4 ArgonFrontierCode (reported by Cognition)47.5%gemini-4-argon_mediummedium, devin-climean score over 5 attemptsCC-BY-4.0epoch
Gemini 4 ArgonFrontierMath Tiers 1-3 (v2)86.0%gemini-4-argon_highhighshare of problems solvedCC-BY-4.0epoch
Gemini 4 ArgonGPQA Diamond96.4%gemini-4-argon_highhighshare of questions correctCC-BY-4.0epoch
Gemini 4 ArgonSimpleQA Verified76.0%gemini-4-argon_highhighshare of questions correctCC-BY-4.0epoch
Gemma 4 31BArena text: business and management1,430Gemma 4 31Bnot statedArena scoreCC-BY-4.0lmarena
Gemma 4 31BArena text: coding1,459Gemma 4 31Bnot statedArena scoreCC-BY-4.0lmarena
Gemma 4 31BArena text: creative writing1,416Gemma 4 31Bnot statedArena scoreCC-BY-4.0lmarena
Gemma 4 31BArena text: following instructions1,434Gemma 4 31Bnot statedArena scoreCC-BY-4.0lmarena
Gemma 4 31BArena text: maths1,464Gemma 4 31Bnot statedArena scoreCC-BY-4.0lmarena
Gemma 4 31BArena text: overall preference1,443Gemma 4 31Bnot statedArena scoreCC-BY-4.0lmarena
Gemma 4 31BArena WebDev: building web apps1,369Gemma 4 31Bnot statedArena scoreCC-BY-4.0lmarena
GLM-4.7Arena text: business and management1,429GLM-4.7not statedArena scoreCC-BY-4.0lmarena
GLM-4.7Arena text: coding1,454GLM-4.7not statedArena scoreCC-BY-4.0lmarena
GLM-4.7Arena text: creative writing1,401GLM-4.7not statedArena scoreCC-BY-4.0lmarena
GLM-4.7Arena text: following instructions1,412GLM-4.7not statedArena scoreCC-BY-4.0lmarena
GLM-4.7Arena text: maths1,424GLM-4.7not statedArena scoreCC-BY-4.0lmarena
GLM-4.7Arena text: overall preference1,435GLM-4.7not statedArena scoreCC-BY-4.0lmarena
GLM-4.7Arena WebDev: building web apps1,435GLM-4.7not statedArena scoreCC-BY-4.0lmarena
GLM-4.7OTIS Mock AIME 2024-202583.3%glm-4.7not statedshare of problems solvedCC-BY-4.0epoch
GLM-4.7GPQA Diamond83.3%glm-4.7not statedshare of questions correctCC-BY-4.0epoch
GLM-4.7SimpleQA Verified32.2%glm-4.7not statedshare of questions correctCC-BY-4.0epoch
GLM-5Arena text: business and management1,430GLM-5not statedArena scoreCC-BY-4.0lmarena
GLM-5Arena text: coding1,461GLM-5not statedArena scoreCC-BY-4.0lmarena
GLM-5Arena text: creative writing1,438GLM-5not statedArena scoreCC-BY-4.0lmarena
GLM-5Arena text: following instructions1,428GLM-5not statedArena scoreCC-BY-4.0lmarena
GLM-5Arena text: maths1,440GLM-5not statedArena scoreCC-BY-4.0lmarena
GLM-5Arena text: overall preference1,446GLM-5not statedArena scoreCC-BY-4.0lmarena
GLM-5Arena WebDev: building web apps1,436GLM-5not statedArena scoreCC-BY-4.0lmarena
GLM-5OTIS Mock AIME 2024-202580.0%glm-5not statedshare of problems solvedCC-BY-4.0epoch
GLM-5GPQA Diamond87.8%glm-5not statedshare of questions correctCC-BY-4.0epoch
GLM-5SWE-bench Verified (Epoch runs)72.1%glm-5not statedshare of issues resolvedCC-BY-4.0epoch
GLM-5.1Arena text: business and management1,453GLM-5.1not statedArena scoreCC-BY-4.0lmarena
GLM-5.1Arena text: coding1,486GLM-5.1not statedArena scoreCC-BY-4.0lmarena
GLM-5.1Arena text: creative writing1,453GLM-5.1not statedArena scoreCC-BY-4.0lmarena
GLM-5.1Arena text: following instructions1,452GLM-5.1not statedArena scoreCC-BY-4.0lmarena
GLM-5.1Arena text: maths1,473GLM-5.1not statedArena scoreCC-BY-4.0lmarena
GLM-5.1Arena text: overall preference1,463GLM-5.1not statedArena scoreCC-BY-4.0lmarena
GLM-5.1Arena WebDev: building web apps1,508GLM-5.1not statedArena scoreCC-BY-4.0lmarena
GLM-5.1OTIS Mock AIME 2024-202593.3%glm-5.1not statedshare of problems solvedCC-BY-4.0epoch
GLM-5.1FrontierMath Tiers 1-3 (v2)24.9%glm-5.1_nonenoneshare of problems solvedCC-BY-4.0epoch
GLM-5.1GPQA Diamond89.9%glm-5.1not statedshare of questions correctCC-BY-4.0epoch
GLM-5.1SimpleQA Verified34.0%glm-5.1not statedshare of questions correctCC-BY-4.0epoch
GLM-5.1SWE-bench Verified (Epoch runs)74.2%glm-5.1not statedshare of issues resolvedCC-BY-4.0epoch
GLM-5.2Arena text: business and management1,451GLM-5.2 (Max)MaxArena scoreCC-BY-4.0lmarena
GLM-5.2Arena text: coding1,487GLM-5.2 (Max)MaxArena scoreCC-BY-4.0lmarena
GLM-5.2Arena text: creative writing1,461GLM-5.2 (Max)MaxArena scoreCC-BY-4.0lmarena
GLM-5.2Arena text: following instructions1,466GLM-5.2 (Max)MaxArena scoreCC-BY-4.0lmarena
GLM-5.2Arena text: maths1,484GLM-5.2 (Max)MaxArena scoreCC-BY-4.0lmarena
GLM-5.2Arena text: overall preference1,470GLM-5.2 (Max)MaxArena scoreCC-BY-4.0lmarena
GLM-5.2Arena WebDev: building web apps1,606GLM-5.2 (Max)MaxArena scoreCC-BY-4.0lmarena
GLM-5.2OTIS Mock AIME 2024-202586.4%glm-5.2_maxmaxshare of problems solvedCC-BY-4.0epoch
GLM-5.2FrontierCode (reported by Cognition)24.5%glm-5.2_nonenone, mini-swe-agentmean score over 5 attemptsCC-BY-4.0epoch
GLM-5.2FrontierMath Tiers 1-3 (v2)59.2%glm-5.2_maxmaxshare of problems solvedCC-BY-4.0epoch
GLM-5.2GPQA Diamond91.9%glm-5.2_maxmaxshare of questions correctCC-BY-4.0epoch
GLM-5.2SimpleQA Verified34.2%glm-5.2_maxmaxshare of questions correctCC-BY-4.0epoch
GLM-5.2SWE-bench Verified (Epoch runs)78.7%glm-5.2_maxmaxshare of issues resolvedCC-BY-4.0epoch
GLM-5.3Arena text: business and management1,461GLM-5.3 (Max)MaxArena scoreCC-BY-4.0lmarena
GLM-5.3Arena text: coding1,497GLM-5.3 (Max)MaxArena scoreCC-BY-4.0lmarena
GLM-5.3Arena text: creative writing1,459GLM-5.3 (Max)MaxArena scoreCC-BY-4.0lmarena
GLM-5.3Arena text: following instructions1,481GLM-5.3 (Max)MaxArena scoreCC-BY-4.0lmarena
GLM-5.3Arena text: maths1,497GLM-5.3 (Max)MaxArena scoreCC-BY-4.0lmarena
GLM-5.3Arena text: overall preference1,474GLM-5.3 (Max)MaxArena scoreCC-BY-4.0lmarena
GLM-5.3Arena WebDev: building web apps1,623GLM-5.3 (Max)MaxArena scoreCC-BY-4.0lmarena
GLM-5.3OTIS Mock AIME 2024-202591.1%glm-5.3_maxmaxshare of problems solvedCC-BY-4.0epoch
GLM-5.3FrontierCode (reported by Cognition)40.1%glm-5.3_maxmax, devin-climean score over 5 attemptsCC-BY-4.0epoch
GLM-5.3FrontierMath Tiers 1-3 (v2)68.8%glm-5.3_maxmaxshare of problems solvedCC-BY-4.0epoch
GLM-5.3GPQA Diamond90.9%glm-5.3_maxmaxshare of questions correctCC-BY-4.0epoch
GLM-5.3SimpleQA Verified41.0%glm-5.3_maxmaxshare of questions correctCC-BY-4.0epoch
GLM-5.3-FlashArena text: business and management1,461GLM-5.3-Flashnot statedArena scoreCC-BY-4.0lmarena
GLM-5.3-FlashArena text: coding1,509GLM-5.3-Flashnot statedArena scoreCC-BY-4.0lmarena
GLM-5.3-FlashArena text: creative writing1,441GLM-5.3-Flashnot statedArena scoreCC-BY-4.0lmarena
GLM-5.3-FlashArena text: following instructions1,481GLM-5.3-Flashnot statedArena scoreCC-BY-4.0lmarena
GLM-5.3-FlashArena text: maths1,502GLM-5.3-Flashnot statedArena scoreCC-BY-4.0lmarena
GLM-5.3-FlashArena text: overall preference1,473GLM-5.3-Flashnot statedArena scoreCC-BY-4.0lmarena
GLM-5.3-FlashArena WebDev: building web apps1,607GLM-5.3-Flashnot statedArena scoreCC-BY-4.0lmarena
GLM-5.3-FlashOTIS Mock AIME 2024-202593.9%glm-5.3-flash_maxmaxshare of problems solvedCC-BY-4.0epoch
GLM-5.3-FlashFrontierCode (reported by Cognition)31.8%glm-5.3-flash_maxmax, devin-climean score over 5 attemptsCC-BY-4.0epoch
GLM-5.3-FlashFrontierMath Tiers 1-3 (v2)55.8%glm-5.3-flash_maxmaxshare of problems solvedCC-BY-4.0epoch
GLM-5.3-FlashGPQA Diamond90.2%glm-5.3-flash_maxmaxshare of questions correctCC-BY-4.0epoch
GPT-5.1Arena text: business and management1,431GPT-5.1 (High)HighArena scoreCC-BY-4.0lmarena
GPT-5.1Arena text: coding1,453GPT-5.1 (High)HighArena scoreCC-BY-4.0lmarena
GPT-5.1Arena text: creative writing1,427GPT-5.1 (High)HighArena scoreCC-BY-4.0lmarena
GPT-5.1Arena text: following instructions1,443GPT-5.1 (High)HighArena scoreCC-BY-4.0lmarena
GPT-5.1Arena text: maths1,446GPT-5.1 (High)HighArena scoreCC-BY-4.0lmarena
GPT-5.1Arena text: overall preference1,443GPT-5.1 (High)HighArena scoreCC-BY-4.0lmarena
GPT-5.1Arena WebDev: building web apps1,396GPT-5.1 (Medium)MediumArena scoreCC-BY-4.0lmarena
GPT-5.4Arena text: business and management1,475GPT-5.4 (High)HighArena scoreCC-BY-4.0lmarena
GPT-5.4Arena text: coding1,496GPT-5.4 (High)HighArena scoreCC-BY-4.0lmarena
GPT-5.4Arena text: creative writing1,439GPT-5.4 (High)HighArena scoreCC-BY-4.0lmarena
GPT-5.4Arena text: following instructions1,469GPT-5.4 (High)HighArena scoreCC-BY-4.0lmarena
GPT-5.4Arena text: maths1,488GPT-5.4 (High)HighArena scoreCC-BY-4.0lmarena
GPT-5.4Arena text: overall preference1,469GPT-5.4 (High)HighArena scoreCC-BY-4.0lmarena
GPT-5.4Arena WebDev: building web apps1,401GPT-5.4not statedArena scoreCC-BY-4.0lmarena
GPT-5.5Arena text: business and management1,478GPT-5.5 (High)HighArena scoreCC-BY-4.0lmarena
GPT-5.5Arena text: coding1,494GPT-5.5 (High)HighArena scoreCC-BY-4.0lmarena
GPT-5.5Arena text: creative writing1,455GPT-5.5 (High)HighArena scoreCC-BY-4.0lmarena
GPT-5.5Arena text: following instructions1,479GPT-5.5 (High)HighArena scoreCC-BY-4.0lmarena
GPT-5.5Arena text: maths1,478GPT-5.5 (High)HighArena scoreCC-BY-4.0lmarena
GPT-5.5Arena text: overall preference1,472GPT-5.5 (High)HighArena scoreCC-BY-4.0lmarena
GPT-5.5OTIS Mock AIME 2024-202584.4%gpt-5.5_lowlowshare of problems solvedCC-BY-4.0epoch
GPT-5.5FrontierCode (reported by Cognition)43.0%gpt-5.5_unknownxhigh, codexmean score over 5 attemptsCC-BY-4.0epoch
GPT-5.5FrontierMath Tiers 1-3 (v2)85.3%gpt-5.5_xhighxhighshare of problems solvedCC-BY-4.0epoch
GPT-5.5GPQA Diamond90.7%gpt-5.5_lowlowshare of questions correctCC-BY-4.0epoch
GPT-5.5SimpleQA Verified63.0%gpt-5.5_xhighxhighshare of questions correctCC-BY-4.0epoch
GPT-5.5 InstantArena text: business and management1,410GPT-5.5 Instantnot statedArena scoreCC-BY-4.0lmarena
GPT-5.5 InstantArena text: coding1,433GPT-5.5 Instantnot statedArena scoreCC-BY-4.0lmarena
GPT-5.5 InstantArena text: creative writing1,418GPT-5.5 Instantnot statedArena scoreCC-BY-4.0lmarena
GPT-5.5 InstantArena text: following instructions1,406GPT-5.5 Instantnot statedArena scoreCC-BY-4.0lmarena
GPT-5.5 InstantArena text: maths1,420GPT-5.5 Instantnot statedArena scoreCC-BY-4.0lmarena
GPT-5.5 InstantArena text: overall preference1,419GPT-5.5 Instantnot statedArena scoreCC-BY-4.0lmarena
GPT-5.5 InstantOTIS Mock AIME 2024-202568.1%gpt-5.5-instantnot statedshare of problems solvedCC-BY-4.0epoch
GPT-5.5 InstantFrontierMath Tiers 1-3 (v2)26.3%gpt-5.5-instantnot statedshare of problems solvedCC-BY-4.0epoch
GPT-5.5 InstantGPQA Diamond82.5%gpt-5.5-instantnot statedshare of questions correctCC-BY-4.0epoch
GPT-5.6 LunaArena text: business and management1,439GPT-5.6 Luna (xHigh)xHighArena scoreCC-BY-4.0lmarena
GPT-5.6 LunaArena text: coding1,466GPT-5.6 Luna (xHigh)xHighArena scoreCC-BY-4.0lmarena
GPT-5.6 LunaArena text: creative writing1,396GPT-5.6 Luna (xHigh)xHighArena scoreCC-BY-4.0lmarena
GPT-5.6 LunaArena text: following instructions1,438GPT-5.6 Luna (xHigh)xHighArena scoreCC-BY-4.0lmarena
GPT-5.6 LunaArena text: maths1,459GPT-5.6 Luna (xHigh)xHighArena scoreCC-BY-4.0lmarena
GPT-5.6 LunaArena text: overall preference1,431GPT-5.6 Luna (xHigh)xHighArena scoreCC-BY-4.0lmarena
GPT-5.6 LunaOTIS Mock AIME 2024-202598.3%gpt-5.6-luna_maxmaxshare of problems solvedCC-BY-4.0epoch
GPT-5.6 LunaFrontierCode (reported by Cognition)39.8%gpt-5.6-luna_unknownmax, codexmean score over 5 attemptsCC-BY-4.0epoch
GPT-5.6 LunaFrontierMath Tiers 1-3 (v2)82.1%gpt-5.6-luna_maxmaxshare of problems solvedCC-BY-4.0epoch
GPT-5.6 LunaGPQA Diamond91.6%gpt-5.6-luna_maxmaxshare of questions correctCC-BY-4.0epoch
GPT-5.6 LunaSimpleQA Verified41.0%gpt-5.6-luna_maxmaxshare of questions correctCC-BY-4.0epoch
GPT-5.6 SolArena text: business and management1,456GPT-5.6 Sol (xHigh)xHighArena scoreCC-BY-4.0lmarena
GPT-5.6 SolArena text: coding1,495GPT-5.6 Sol (xHigh)xHighArena scoreCC-BY-4.0lmarena
GPT-5.6 SolArena text: creative writing1,450GPT-5.6 Sol (xHigh)xHighArena scoreCC-BY-4.0lmarena
GPT-5.6 SolArena text: following instructions1,483GPT-5.6 Sol (xHigh)xHighArena scoreCC-BY-4.0lmarena
GPT-5.6 SolArena text: maths1,472GPT-5.6 Sol (xHigh)xHighArena scoreCC-BY-4.0lmarena
GPT-5.6 SolArena text: overall preference1,457GPT-5.6 Sol (xHigh)xHighArena scoreCC-BY-4.0lmarena
GPT-5.6 SolOTIS Mock AIME 2024-2025100.0%gpt-5.6-sol_maxmaxshare of problems solvedCC-BY-4.0epoch
GPT-5.6 SolFrontierCode (reported by Cognition)47.5%gpt-5.6-sol_unknownmax, codexmean score over 5 attemptsCC-BY-4.0epoch
GPT-5.6 SolFrontierMath Tiers 1-3 (v2)89.1%gpt-5.6-sol_maxmaxshare of problems solvedCC-BY-4.0epoch
GPT-5.6 SolGPQA Diamond93.5%gpt-5.6-sol_maxmaxshare of questions correctCC-BY-4.0epoch
GPT-5.6 SolSimpleQA Verified69.7%gpt-5.6-sol_maxmaxshare of questions correctCC-BY-4.0epoch
GPT-5.6 TerraArena text: business and management1,447GPT-5.6 Terra (xHigh)xHighArena scoreCC-BY-4.0lmarena
GPT-5.6 TerraArena text: coding1,485GPT-5.6 Terra (xHigh)xHighArena scoreCC-BY-4.0lmarena
GPT-5.6 TerraArena text: creative writing1,413GPT-5.6 Terra (xHigh)xHighArena scoreCC-BY-4.0lmarena
GPT-5.6 TerraArena text: following instructions1,455GPT-5.6 Terra (xHigh)xHighArena scoreCC-BY-4.0lmarena
GPT-5.6 TerraArena text: maths1,464GPT-5.6 Terra (xHigh)xHighArena scoreCC-BY-4.0lmarena
GPT-5.6 TerraArena text: overall preference1,448GPT-5.6 Terra (xHigh)xHighArena scoreCC-BY-4.0lmarena
GPT-5.6 TerraOTIS Mock AIME 2024-202599.7%gpt-5.6-terra_maxmaxshare of problems solvedCC-BY-4.0epoch
GPT-5.6 TerraFrontierCode (reported by Cognition)41.3%gpt-5.6-terra_unknownmax, codexmean score over 5 attemptsCC-BY-4.0epoch
GPT-5.6 TerraFrontierMath Tiers 1-3 (v2)86.0%gpt-5.6-terra_maxmaxshare of problems solvedCC-BY-4.0epoch
GPT-5.6 TerraGPQA Diamond93.3%gpt-5.6-terra_maxmaxshare of questions correctCC-BY-4.0epoch
GPT-5.6 TerraSimpleQA Verified43.2%gpt-5.6-terra_maxmaxshare of questions correctCC-BY-4.0epoch
GPT-6.1 SolArena text: business and management1,449GPT-6.1 Sol (Max)MaxArena scoreCC-BY-4.0lmarena
GPT-6.1 SolArena text: coding1,486GPT-6.1 Sol (Max)MaxArena scoreCC-BY-4.0lmarena
GPT-6.1 SolArena text: creative writing1,425GPT-6.1 Sol (Max)MaxArena scoreCC-BY-4.0lmarena
GPT-6.1 SolArena text: following instructions1,467GPT-6.1 Sol (Max)MaxArena scoreCC-BY-4.0lmarena
GPT-6.1 SolArena text: maths1,486GPT-6.1 Sol (Max)MaxArena scoreCC-BY-4.0lmarena
GPT-6.1 SolArena text: overall preference1,446GPT-6.1 Sol (Max)MaxArena scoreCC-BY-4.0lmarena
GPT-6.1 SolArena WebDev: building web apps1,751GPT-6.1 Sol (Max)MaxArena scoreCC-BY-4.0lmarena
GPT-6.1 SolOTIS Mock AIME 2024-2025100.0%gpt-6.1-sol_maxmaxshare of problems solvedCC-BY-4.0epoch
GPT-6.1 SolFrontierCode (reported by Cognition)50.2%gpt-6.1-sol_mediummedium, codexmean score over 5 attemptsCC-BY-4.0epoch
GPT-6.1 SolFrontierMath Tiers 1-3 (v2)93.7%gpt-6.1-sol_maxmaxshare of problems solvedCC-BY-4.0epoch
GPT-6.1 SolGPQA Diamond95.4%gpt-6.1-sol_maxmaxshare of questions correctCC-BY-4.0epoch
GPT-6.1 SolSimpleQA Verified73.9%gpt-6.1-sol_maxmaxshare of questions correctCC-BY-4.0epoch
GPT-6 AstraArena text: business and management1,438GPT-6 Astra (Max)MaxArena scoreCC-BY-4.0lmarena
GPT-6 AstraArena text: coding1,488GPT-6 Astra (Max)MaxArena scoreCC-BY-4.0lmarena
GPT-6 AstraArena text: creative writing1,420GPT-6 Astra (Max)MaxArena scoreCC-BY-4.0lmarena
GPT-6 AstraArena text: following instructions1,454GPT-6 Astra (Max)MaxArena scoreCC-BY-4.0lmarena
GPT-6 AstraArena text: maths1,468GPT-6 Astra (Max)MaxArena scoreCC-BY-4.0lmarena
GPT-6 AstraArena text: overall preference1,441GPT-6 Astra (Max)MaxArena scoreCC-BY-4.0lmarena
GPT-6 AstraArena WebDev: building web apps1,773GPT-6 Astra (Max)MaxArena scoreCC-BY-4.0lmarena
GPT-6 AstraOTIS Mock AIME 2024-2025100.0%gpt-6-astra_maxmaxshare of problems solvedCC-BY-4.0epoch
GPT-6 AstraFrontierCode (reported by Cognition)53.3%gpt-6-astra_maxmax, codexmean score over 5 attemptsCC-BY-4.0epoch
GPT-6 AstraFrontierMath Tiers 1-3 (v2)93.7%gpt-6-astra_maxmaxshare of problems solvedCC-BY-4.0epoch
GPT-6 AstraGPQA Diamond95.8%gpt-6-astra_maxmaxshare of questions correctCC-BY-4.0epoch
GPT-6 AstraSimpleQA Verified75.6%gpt-6-astra_maxmaxshare of questions correctCC-BY-4.0epoch
GPT-6 LunaArena text: business and management1,384GPT-6 Luna (Max)MaxArena scoreCC-BY-4.0lmarena
GPT-6 LunaArena text: coding1,437GPT-6 Luna (Max)MaxArena scoreCC-BY-4.0lmarena
GPT-6 LunaArena text: creative writing1,362GPT-6 Luna (Max)MaxArena scoreCC-BY-4.0lmarena
GPT-6 LunaArena text: following instructions1,406GPT-6 Luna (Max)MaxArena scoreCC-BY-4.0lmarena
GPT-6 LunaArena text: maths1,410GPT-6 Luna (Max)MaxArena scoreCC-BY-4.0lmarena
GPT-6 LunaArena text: overall preference1,389GPT-6 Luna (Max)MaxArena scoreCC-BY-4.0lmarena
GPT-6 LunaArena WebDev: building web apps1,579GPT-6 Luna (Max)MaxArena scoreCC-BY-4.0lmarena
GPT-6 LunaOTIS Mock AIME 2024-202598.9%gpt-6-luna_maxmaxshare of problems solvedCC-BY-4.0epoch
GPT-6 LunaFrontierCode (reported by Cognition)42.4%gpt-6-luna_maxmax, codexmean score over 5 attemptsCC-BY-4.0epoch
GPT-6 LunaFrontierMath Tiers 1-3 (v2)78.9%gpt-6-luna_maxmaxshare of problems solvedCC-BY-4.0epoch
GPT-6 LunaGPQA Diamond90.5%gpt-6-luna_maxmaxshare of questions correctCC-BY-4.0epoch
GPT-6 LunaSimpleQA Verified41.4%gpt-6-luna_maxmaxshare of questions correctCC-BY-4.0epoch
GPT-6 SolArena text: business and management1,381GPT-6 Sol (Max)MaxArena scoreCC-BY-4.0lmarena
GPT-6 SolArena text: coding1,447GPT-6 Sol (Max)MaxArena scoreCC-BY-4.0lmarena
GPT-6 SolArena text: creative writing1,378GPT-6 Sol (Max)MaxArena scoreCC-BY-4.0lmarena
GPT-6 SolArena text: following instructions1,411GPT-6 Sol (Max)MaxArena scoreCC-BY-4.0lmarena
GPT-6 SolArena text: maths1,404GPT-6 Sol (Max)MaxArena scoreCC-BY-4.0lmarena
GPT-6 SolArena text: overall preference1,397GPT-6 Sol (Max)MaxArena scoreCC-BY-4.0lmarena
GPT-6 SolArena WebDev: building web apps1,678GPT-6 Sol (Max)MaxArena scoreCC-BY-4.0lmarena
GPT-6 SolOTIS Mock AIME 2024-2025100.0%gpt-6-sol_maxmaxshare of problems solvedCC-BY-4.0epoch
GPT-6 SolFrontierCode (reported by Cognition)49.3%gpt-6-sol_maxmax, codexmean score over 5 attemptsCC-BY-4.0epoch
GPT-6 SolFrontierMath Tiers 1-3 (v2)89.8%gpt-6-sol_maxmaxshare of problems solvedCC-BY-4.0epoch
GPT-6 SolGPQA Diamond94.3%gpt-6-sol_maxmaxshare of questions correctCC-BY-4.0epoch
GPT-6 SolSimpleQA Verified60.7%gpt-6-sol_maxmaxshare of questions correctCC-BY-4.0epoch
Grok 4.20 Beta 1Arena text: business and management1,425Grok 4.20 Beta 1not statedArena scoreCC-BY-4.0lmarena
Grok 4.20 Beta 1Arena text: coding1,447Grok 4.20 Beta 1not statedArena scoreCC-BY-4.0lmarena
Grok 4.20 Beta 1Arena text: creative writing1,438Grok 4.20 Beta 1not statedArena scoreCC-BY-4.0lmarena
Grok 4.20 Beta 1Arena text: following instructions1,415Grok 4.20 Beta 1not statedArena scoreCC-BY-4.0lmarena
Grok 4.20 Beta 1Arena text: maths1,432Grok 4.20 Beta 1not statedArena scoreCC-BY-4.0lmarena
Grok 4.20 Beta 1Arena text: overall preference1,444Grok 4.20 Beta 1not statedArena scoreCC-BY-4.0lmarena
Grok 4.20 Beta (Reasoning, 0309)Arena text: business and management1,432Grok 4.20 Beta (Reasoning, 0309)Reasoning, 0309Arena scoreCC-BY-4.0lmarena
Grok 4.20 Beta (Reasoning, 0309)Arena text: coding1,459Grok 4.20 Beta (Reasoning, 0309)Reasoning, 0309Arena scoreCC-BY-4.0lmarena
Grok 4.20 Beta (Reasoning, 0309)Arena text: creative writing1,432Grok 4.20 Beta (Reasoning, 0309)Reasoning, 0309Arena scoreCC-BY-4.0lmarena
Grok 4.20 Beta (Reasoning, 0309)Arena text: following instructions1,421Grok 4.20 Beta (Reasoning, 0309)Reasoning, 0309Arena scoreCC-BY-4.0lmarena
Grok 4.20 Beta (Reasoning, 0309)Arena text: maths1,455Grok 4.20 Beta (Reasoning, 0309)Reasoning, 0309Arena scoreCC-BY-4.0lmarena
Grok 4.20 Beta (Reasoning, 0309)Arena text: overall preference1,451Grok 4.20 Beta (Reasoning, 0309)Reasoning, 0309Arena scoreCC-BY-4.0lmarena
Grok 4.20 Beta (Reasoning, 0309)Arena WebDev: building web apps1,375Grok 4.20 Beta (Reasoning, 0309)Reasoning, 0309Arena scoreCC-BY-4.0lmarena
Grok 4.20 Multi-Agent Beta (0309)Arena text: business and management1,423Grok 4.20 Multi-Agent Beta (0309)0309Arena scoreCC-BY-4.0lmarena
Grok 4.20 Multi-Agent Beta (0309)Arena text: coding1,457Grok 4.20 Multi-Agent Beta (0309)0309Arena scoreCC-BY-4.0lmarena
Grok 4.20 Multi-Agent Beta (0309)Arena text: creative writing1,436Grok 4.20 Multi-Agent Beta (0309)0309Arena scoreCC-BY-4.0lmarena
Grok 4.20 Multi-Agent Beta (0309)Arena text: following instructions1,420Grok 4.20 Multi-Agent Beta (0309)0309Arena scoreCC-BY-4.0lmarena
Grok 4.20 Multi-Agent Beta (0309)Arena text: maths1,441Grok 4.20 Multi-Agent Beta (0309)0309Arena scoreCC-BY-4.0lmarena
Grok 4.20 Multi-Agent Beta (0309)Arena text: overall preference1,451Grok 4.20 Multi-Agent Beta (0309)0309Arena scoreCC-BY-4.0lmarena
Grok 4.3Arena text: business and management1,381Grok 4.3not statedArena scoreCC-BY-4.0lmarena
Grok 4.3Arena text: coding1,414Grok 4.3not statedArena scoreCC-BY-4.0lmarena
Grok 4.3Arena text: creative writing1,380Grok 4.3not statedArena scoreCC-BY-4.0lmarena
Grok 4.3Arena text: following instructions1,366Grok 4.3not statedArena scoreCC-BY-4.0lmarena
Grok 4.3Arena text: maths1,387Grok 4.3not statedArena scoreCC-BY-4.0lmarena
Grok 4.3Arena text: overall preference1,397Grok 4.3not statedArena scoreCC-BY-4.0lmarena
Grok 4.3Arena WebDev: building web apps1,358Grok 4.3not statedArena scoreCC-BY-4.0lmarena
Grok 4.5Arena text: business and management1,441Grok 4.5not statedArena scoreCC-BY-4.0lmarena
Grok 4.5Arena text: coding1,474Grok 4.5not statedArena scoreCC-BY-4.0lmarena
Grok 4.5Arena text: creative writing1,442Grok 4.5not statedArena scoreCC-BY-4.0lmarena
Grok 4.5Arena text: following instructions1,445Grok 4.5not statedArena scoreCC-BY-4.0lmarena
Grok 4.5Arena text: maths1,459Grok 4.5not statedArena scoreCC-BY-4.0lmarena
Grok 4.5Arena text: overall preference1,448Grok 4.5not statedArena scoreCC-BY-4.0lmarena
Grok 4.5Arena WebDev: building web apps1,553Grok 4.5not statedArena scoreCC-BY-4.0lmarena
Grok 4.5OTIS Mock AIME 2024-202597.8%grok-4.5_highhighshare of problems solvedCC-BY-4.0epoch
Grok 4.5FrontierCode (reported by Cognition)42.4%grok-4.5_unknownhigh, grok-buildmean score over 5 attemptsCC-BY-4.0epoch
Grok 4.5FrontierMath Tiers 1-3 (v2)57.2%grok-4.5_highhighshare of problems solvedCC-BY-4.0epoch
Grok 4.5GPQA Diamond93.4%grok-4.5_highhighshare of questions correctCC-BY-4.0epoch
Grok 4.5SimpleQA Verified48.3%grok-4.5_highhighshare of questions correctCC-BY-4.0epoch
Grok 4.6Arena text: business and management1,416Grok 4.6 (High)HighArena scoreCC-BY-4.0lmarena
Grok 4.6Arena text: coding1,464Grok 4.6 (High)HighArena scoreCC-BY-4.0lmarena
Grok 4.6Arena text: creative writing1,426Grok 4.6 (High)HighArena scoreCC-BY-4.0lmarena
Grok 4.6Arena text: following instructions1,431Grok 4.6 (High)HighArena scoreCC-BY-4.0lmarena
Grok 4.6Arena text: maths1,415Grok 4.6 (High)HighArena scoreCC-BY-4.0lmarena
Grok 4.6Arena text: overall preference1,429Grok 4.6 (High)HighArena scoreCC-BY-4.0lmarena
Grok 4.6Arena WebDev: building web apps1,616Grok 4.6 (High)HighArena scoreCC-BY-4.0lmarena
Grok 4.6OTIS Mock AIME 2024-202599.2%grok-4.6_xhighxhighshare of problems solvedCC-BY-4.0epoch
Grok 4.6FrontierCode (reported by Cognition)48.0%grok-4.6_unknownhigh, grok-buildmean score over 5 attemptsCC-BY-4.0epoch
Grok 4.6FrontierMath Tiers 1-3 (v2)66.0%grok-4.6_xhighxhighshare of problems solvedCC-BY-4.0epoch
Grok 4.6GPQA Diamond93.2%grok-4.6_xhighxhighshare of questions correctCC-BY-4.0epoch
Grok 4.6SimpleQA Verified48.9%grok-4.6_xhighxhighshare of questions correctCC-BY-4.0epoch
Grok 4.7Arena text: business and management1,390Grok 4.7 (xHigh)xHighArena scoreCC-BY-4.0lmarena
Grok 4.7Arena text: coding1,428Grok 4.7 (xHigh)xHighArena scoreCC-BY-4.0lmarena
Grok 4.7Arena text: creative writing1,381Grok 4.7 (xHigh)xHighArena scoreCC-BY-4.0lmarena
Grok 4.7Arena text: following instructions1,406Grok 4.7 (xHigh)xHighArena scoreCC-BY-4.0lmarena
Grok 4.7Arena text: maths1,413Grok 4.7 (xHigh)xHighArena scoreCC-BY-4.0lmarena
Grok 4.7Arena text: overall preference1,398Grok 4.7 (xHigh)xHighArena scoreCC-BY-4.0lmarena
Grok 4.7Arena WebDev: building web apps1,637Grok 4.7 (xHigh)xHighArena scoreCC-BY-4.0lmarena
Grok 4.7OTIS Mock AIME 2024-202598.1%grok-4.7_xhighxhighshare of problems solvedCC-BY-4.0epoch
Grok 4.7FrontierCode (reported by Cognition)47.6%grok-4.7_unknownhigh, grok-buildmean score over 5 attemptsCC-BY-4.0epoch
Grok 4.7FrontierMath Tiers 1-3 (v2)53.0%grok-4.7_xhighxhighshare of problems solvedCC-BY-4.0epoch
Grok 4.7GPQA Diamond92.7%grok-4.7_xhighxhighshare of questions correctCC-BY-4.0epoch
Grok 4.7SimpleQA Verified56.0%grok-4.7_xhighxhighshare of questions correctCC-BY-4.0epoch
Inkling-SmallArena text: business and management1,414Inkling-Smallnot statedArena scoreCC-BY-4.0lmarena
Inkling-SmallArena text: coding1,451Inkling-Smallnot statedArena scoreCC-BY-4.0lmarena
Inkling-SmallArena text: creative writing1,336Inkling-Smallnot statedArena scoreCC-BY-4.0lmarena
Inkling-SmallArena text: following instructions1,401Inkling-Smallnot statedArena scoreCC-BY-4.0lmarena
Inkling-SmallArena text: maths1,458Inkling-Smallnot statedArena scoreCC-BY-4.0lmarena
Inkling-SmallArena text: overall preference1,414Inkling-Smallnot statedArena scoreCC-BY-4.0lmarena
Inkling-SmallArena WebDev: building web apps1,409Inkling-Smallnot statedArena scoreCC-BY-4.0lmarena
Inkling-SmallOTIS Mock AIME 2024-202590.0%inkling-small_xhighxhighshare of problems solvedCC-BY-4.0epoch
Inkling-SmallFrontierMath Tiers 1-3 (v2)46.3%inkling-small_xhighxhighshare of problems solvedCC-BY-4.0epoch
Inkling-SmallGPQA Diamond88.5%inkling-small_xhighxhighshare of questions correctCC-BY-4.0epoch
Inkling-SmallSimpleQA Verified19.1%inkling-small_xhighxhighshare of questions correctCC-BY-4.0epoch
Kimi K2.5Arena text: business and management1,427Kimi K2.5 (Thinking)ThinkingArena scoreCC-BY-4.0lmarena
Kimi K2.5Arena text: coding1,474Kimi K2.5 (Thinking)ThinkingArena scoreCC-BY-4.0lmarena
Kimi K2.5Arena text: creative writing1,422Kimi K2.5 (Thinking)ThinkingArena scoreCC-BY-4.0lmarena
Kimi K2.5Arena text: following instructions1,431Kimi K2.5 (Thinking)ThinkingArena scoreCC-BY-4.0lmarena
Kimi K2.5Arena text: maths1,470Kimi K2.5 (Thinking)ThinkingArena scoreCC-BY-4.0lmarena
Kimi K2.5Arena text: overall preference1,445Kimi K2.5 (Thinking)ThinkingArena scoreCC-BY-4.0lmarena
Kimi K2.5Arena WebDev: building web apps1,436Kimi K2.5 (Thinking)ThinkingArena scoreCC-BY-4.0lmarena
Kimi K2.5SimpleQA Verified34.3%kimi-k2.5not statedshare of questions correctCC-BY-4.0epoch
Kimi K2.5SWE-bench Verified (Epoch runs)73.8%kimi-k2.5not statedshare of issues resolvedCC-BY-4.0epoch
Kimi K2.6Arena text: business and management1,445Kimi K2.6not statedArena scoreCC-BY-4.0lmarena
Kimi K2.6Arena text: coding1,489Kimi K2.6not statedArena scoreCC-BY-4.0lmarena
Kimi K2.6Arena text: creative writing1,433Kimi K2.6not statedArena scoreCC-BY-4.0lmarena
Kimi K2.6Arena text: following instructions1,451Kimi K2.6not statedArena scoreCC-BY-4.0lmarena
Kimi K2.6Arena text: maths1,474Kimi K2.6not statedArena scoreCC-BY-4.0lmarena
Kimi K2.6Arena text: overall preference1,456Kimi K2.6not statedArena scoreCC-BY-4.0lmarena
Kimi K2.6Arena WebDev: building web apps1,509Kimi K2.6not statedArena scoreCC-BY-4.0lmarena
Kimi K2.6OTIS Mock AIME 2024-202596.1%kimi-k2.6not statedshare of problems solvedCC-BY-4.0epoch
Kimi K2.6FrontierMath Tiers 1-3 (v2)57.2%kimi-k2.6not statedshare of problems solvedCC-BY-4.0epoch
Kimi K2.6GPQA Diamond90.8%kimi-k2.6not statedshare of questions correctCC-BY-4.0epoch
Kimi K2.6SimpleQA Verified34.9%kimi-k2.6not statedshare of questions correctCC-BY-4.0epoch
Kimi K2.6SWE-bench Verified (Epoch runs)76.7%kimi-k2.6not statedshare of issues resolvedCC-BY-4.0epoch
Kimi K3Arena text: business and management1,469Kimi K3 (Max)MaxArena scoreCC-BY-4.0lmarena
Kimi K3Arena text: coding1,508Kimi K3 (Max)MaxArena scoreCC-BY-4.0lmarena
Kimi K3Arena text: creative writing1,457Kimi K3 (Max)MaxArena scoreCC-BY-4.0lmarena
Kimi K3Arena text: following instructions1,485Kimi K3 (Max)MaxArena scoreCC-BY-4.0lmarena
Kimi K3Arena text: maths1,492Kimi K3 (Max)MaxArena scoreCC-BY-4.0lmarena
Kimi K3Arena text: overall preference1,476Kimi K3 (Max)MaxArena scoreCC-BY-4.0lmarena
Kimi K3Arena WebDev: building web apps1,656Kimi K3 (Max)MaxArena scoreCC-BY-4.0lmarena
Kimi K3OTIS Mock AIME 2024-202597.2%kimi-k3_maxmaxshare of problems solvedCC-BY-4.0epoch
Kimi K3FrontierCode (reported by Cognition)44.2%kimi-k3_unknownnone, mini-swe-agentmean score over 5 attemptsCC-BY-4.0epoch
Kimi K3FrontierMath Tiers 1-3 (v2)72.2%kimi-k3_maxmaxshare of problems solvedCC-BY-4.0epoch
Kimi K3GPQA Diamond93.1%kimi-k3_maxmaxshare of questions correctCC-BY-4.0epoch
Kimi K3SimpleQA Verified50.6%kimi-k3_maxmaxshare of questions correctCC-BY-4.0epoch
MiMo-V2.5-ProArena text: business and management1,467MiMo-V2.5-Pronot statedArena scoreCC-BY-4.0lmarena
MiMo-V2.5-ProArena text: coding1,501MiMo-V2.5-Pronot statedArena scoreCC-BY-4.0lmarena
MiMo-V2.5-ProArena text: creative writing1,438MiMo-V2.5-Pronot statedArena scoreCC-BY-4.0lmarena
MiMo-V2.5-ProArena text: following instructions1,475MiMo-V2.5-Pronot statedArena scoreCC-BY-4.0lmarena
MiMo-V2.5-ProArena text: maths1,481MiMo-V2.5-Pronot statedArena scoreCC-BY-4.0lmarena
MiMo-V2.5-ProArena text: overall preference1,464MiMo-V2.5-Pronot statedArena scoreCC-BY-4.0lmarena
MiMo-V2.5-ProArena WebDev: building web apps1,480MiMo-V2.5-Pronot statedArena scoreCC-BY-4.0lmarena
MiMo-V2.6-FlashArena text: business and management1,450MiMo-V2.6-Flashnot statedArena scoreCC-BY-4.0lmarena
MiMo-V2.6-FlashArena text: coding1,498MiMo-V2.6-Flashnot statedArena scoreCC-BY-4.0lmarena
MiMo-V2.6-FlashArena text: creative writing1,405MiMo-V2.6-Flashnot statedArena scoreCC-BY-4.0lmarena
MiMo-V2.6-FlashArena text: following instructions1,449MiMo-V2.6-Flashnot statedArena scoreCC-BY-4.0lmarena
MiMo-V2.6-FlashArena text: maths1,468MiMo-V2.6-Flashnot statedArena scoreCC-BY-4.0lmarena
MiMo-V2.6-FlashArena text: overall preference1,451MiMo-V2.6-Flashnot statedArena scoreCC-BY-4.0lmarena
MiMo-V2.6-FlashArena WebDev: building web apps1,637MiMo-V2.6-Flashnot statedArena scoreCC-BY-4.0lmarena
MiMo-V2.6-ProArena text: business and management1,468MiMo-V2.6-Pronot statedArena scoreCC-BY-4.0lmarena
MiMo-V2.6-ProArena text: coding1,534MiMo-V2.6-Pronot statedArena scoreCC-BY-4.0lmarena
MiMo-V2.6-ProArena text: creative writing1,464MiMo-V2.6-Pronot statedArena scoreCC-BY-4.0lmarena
MiMo-V2.6-ProArena text: following instructions1,492MiMo-V2.6-Pronot statedArena scoreCC-BY-4.0lmarena
MiMo-V2.6-ProArena text: maths1,501MiMo-V2.6-Pronot statedArena scoreCC-BY-4.0lmarena
MiMo-V2.6-ProArena text: overall preference1,491MiMo-V2.6-Pronot statedArena scoreCC-BY-4.0lmarena
MiMo-V2.6-ProArena WebDev: building web apps1,625MiMo-V2.6-Pronot statedArena scoreCC-BY-4.0lmarena
Mistral Large 4Arena text: business and management1,439Mistral Large 4not statedArena scoreCC-BY-4.0lmarena
Mistral Large 4Arena text: coding1,478Mistral Large 4not statedArena scoreCC-BY-4.0lmarena
Mistral Large 4Arena text: creative writing1,361Mistral Large 4not statedArena scoreCC-BY-4.0lmarena
Mistral Large 4Arena text: following instructions1,419Mistral Large 4not statedArena scoreCC-BY-4.0lmarena
Mistral Large 4Arena text: maths1,468Mistral Large 4not statedArena scoreCC-BY-4.0lmarena
Mistral Large 4Arena text: overall preference1,430Mistral Large 4not statedArena scoreCC-BY-4.0lmarena
Mistral Large 4Arena WebDev: building web apps1,544Mistral Large 4not statedArena scoreCC-BY-4.0lmarena
Mistral Large 4OTIS Mock AIME 2024-202598.3%mistral-large-4-0_highhighshare of problems solvedCC-BY-4.0epoch
Mistral Large 4FrontierCode (reported by Cognition)39.4%mistral-large-4-0_highhigh, chiselmean score over 5 attemptsCC-BY-4.0epoch
Mistral Large 4FrontierMath Tiers 1-3 (v2)58.6%mistral-large-4-0_highhighshare of problems solvedCC-BY-4.0epoch
Mistral Large 4GPQA Diamond88.1%mistral-large-4-0_highhighshare of questions correctCC-BY-4.0epoch
Mistral Large 4SimpleQA Verified20.0%mistral-large-4-0_highhighshare of questions correctCC-BY-4.0epoch
Muse SparkArena text: business and management1,462Muse Sparknot statedArena scoreCC-BY-4.0lmarena
Muse SparkArena text: coding1,481Muse Sparknot statedArena scoreCC-BY-4.0lmarena
Muse SparkArena text: creative writing1,459Muse Sparknot statedArena scoreCC-BY-4.0lmarena
Muse SparkArena text: following instructions1,442Muse Sparknot statedArena scoreCC-BY-4.0lmarena
Muse SparkArena text: maths1,456Muse Sparknot statedArena scoreCC-BY-4.0lmarena
Muse SparkArena text: overall preference1,474Muse Sparknot statedArena scoreCC-BY-4.0lmarena
Muse SparkOTIS Mock AIME 2024-202588.9%muse-sparknot statedshare of problems solvedCC-BY-4.0epoch
Muse SparkGPQA Diamond89.8%muse-sparknot statedshare of questions correctCC-BY-4.0epoch
Muse Spark 1.1Arena text: business and management1,470Muse Spark 1.1not statedArena scoreCC-BY-4.0lmarena
Muse Spark 1.1Arena text: coding1,498Muse Spark 1.1not statedArena scoreCC-BY-4.0lmarena
Muse Spark 1.1Arena text: creative writing1,435Muse Spark 1.1not statedArena scoreCC-BY-4.0lmarena
Muse Spark 1.1Arena text: following instructions1,457Muse Spark 1.1not statedArena scoreCC-BY-4.0lmarena
Muse Spark 1.1Arena text: maths1,482Muse Spark 1.1not statedArena scoreCC-BY-4.0lmarena
Muse Spark 1.1Arena text: overall preference1,478Muse Spark 1.1not statedArena scoreCC-BY-4.0lmarena
Muse Spark 1.1Arena WebDev: building web apps1,542Muse Spark 1.1not statedArena scoreCC-BY-4.0lmarena
Muse Spark 1.1SimpleQA Verified57.8%muse-spark-1.1not statedshare of questions correctCC-BY-4.0epoch
Muse Spark 1.2Arena text: business and management1,484Muse Spark 1.2 (xHigh)xHighArena scoreCC-BY-4.0lmarena
Muse Spark 1.2Arena text: coding1,506Muse Spark 1.2 (xHigh)xHighArena scoreCC-BY-4.0lmarena
Muse Spark 1.2Arena text: creative writing1,437Muse Spark 1.2 (xHigh)xHighArena scoreCC-BY-4.0lmarena
Muse Spark 1.2Arena text: following instructions1,465Muse Spark 1.2 (xHigh)xHighArena scoreCC-BY-4.0lmarena
Muse Spark 1.2Arena text: maths1,485Muse Spark 1.2 (xHigh)xHighArena scoreCC-BY-4.0lmarena
Muse Spark 1.2Arena text: overall preference1,483Muse Spark 1.2 (xHigh)xHighArena scoreCC-BY-4.0lmarena
Muse Spark 1.2Arena WebDev: building web apps1,534Muse Spark 1.2 (xHigh)xHighArena scoreCC-BY-4.0lmarena
Muse Spark 1.2SimpleQA Verified60.3%muse-spark-1.2_xhighxhighshare of questions correctCC-BY-4.0epoch
Muse Spark 1.3Arena text: business and management1,486Muse Spark 1.3 (Max)MaxArena scoreCC-BY-4.0lmarena
Muse Spark 1.3Arena text: coding1,509Muse Spark 1.3 (Max)MaxArena scoreCC-BY-4.0lmarena
Muse Spark 1.3Arena text: creative writing1,464Muse Spark 1.3 (Max)MaxArena scoreCC-BY-4.0lmarena
Muse Spark 1.3Arena text: following instructions1,478Muse Spark 1.3 (Max)MaxArena scoreCC-BY-4.0lmarena
Muse Spark 1.3Arena text: maths1,501Muse Spark 1.3 (Max)MaxArena scoreCC-BY-4.0lmarena
Muse Spark 1.3Arena text: overall preference1,491Muse Spark 1.3 (Max)MaxArena scoreCC-BY-4.0lmarena
Muse Spark 1.3Arena WebDev: building web apps1,656Muse Spark 1.3 (Max)MaxArena scoreCC-BY-4.0lmarena
Muse Spark 1.3OTIS Mock AIME 2024-202599.2%muse-spark-1.3_xhighxhighshare of problems solvedCC-BY-4.0epoch
Muse Spark 1.3FrontierMath Tiers 1-3 (v2)74.4%muse-spark-1.3_xhighxhighshare of problems solvedCC-BY-4.0epoch
Muse WonderArena text: business and management1,487Muse Wonder (Max)MaxArena scoreCC-BY-4.0lmarena
Muse WonderArena text: coding1,518Muse Wonder (Max)MaxArena scoreCC-BY-4.0lmarena
Muse WonderArena text: creative writing1,488Muse Wonder (Max)MaxArena scoreCC-BY-4.0lmarena
Muse WonderArena text: following instructions1,494Muse Wonder (Max)MaxArena scoreCC-BY-4.0lmarena
Muse WonderArena text: maths1,477Muse Wonder (Max)MaxArena scoreCC-BY-4.0lmarena
Muse WonderArena text: overall preference1,499Muse Wonder (Max)MaxArena scoreCC-BY-4.0lmarena
Muse WonderArena WebDev: building web apps1,652Muse Wonder (Max)MaxArena scoreCC-BY-4.0lmarena
NVIDIA-Nemotron-3-Ultra-550B-A55B (NVFP4)Arena text: business and management1,435NVIDIA-Nemotron-3-Ultra-550B-A55B (NVFP4)NVFP4Arena scoreCC-BY-4.0lmarena
NVIDIA-Nemotron-3-Ultra-550B-A55B (NVFP4)Arena text: coding1,468NVIDIA-Nemotron-3-Ultra-550B-A55B (NVFP4)NVFP4Arena scoreCC-BY-4.0lmarena
NVIDIA-Nemotron-3-Ultra-550B-A55B (NVFP4)Arena text: creative writing1,398NVIDIA-Nemotron-3-Ultra-550B-A55B (NVFP4)NVFP4Arena scoreCC-BY-4.0lmarena
NVIDIA-Nemotron-3-Ultra-550B-A55B (NVFP4)Arena text: following instructions1,418NVIDIA-Nemotron-3-Ultra-550B-A55B (NVFP4)NVFP4Arena scoreCC-BY-4.0lmarena
NVIDIA-Nemotron-3-Ultra-550B-A55B (NVFP4)Arena text: maths1,458NVIDIA-Nemotron-3-Ultra-550B-A55B (NVFP4)NVFP4Arena scoreCC-BY-4.0lmarena
NVIDIA-Nemotron-3-Ultra-550B-A55B (NVFP4)Arena text: overall preference1,445NVIDIA-Nemotron-3-Ultra-550B-A55B (NVFP4)NVFP4Arena scoreCC-BY-4.0lmarena
Qwen3.5-397B-A17BArena text: business and management1,435Qwen3.5-397B-A17Bnot statedArena scoreCC-BY-4.0lmarena
Qwen3.5-397B-A17BArena text: coding1,464Qwen3.5-397B-A17Bnot statedArena scoreCC-BY-4.0lmarena
Qwen3.5-397B-A17BArena text: creative writing1,402Qwen3.5-397B-A17Bnot statedArena scoreCC-BY-4.0lmarena
Qwen3.5-397B-A17BArena text: following instructions1,423Qwen3.5-397B-A17Bnot statedArena scoreCC-BY-4.0lmarena
Qwen3.5-397B-A17BArena text: maths1,454Qwen3.5-397B-A17Bnot statedArena scoreCC-BY-4.0lmarena
Qwen3.5-397B-A17BArena text: overall preference1,438Qwen3.5-397B-A17Bnot statedArena scoreCC-BY-4.0lmarena
Qwen3.5-397B-A17BArena WebDev: building web apps1,400Qwen3.5-397B-A17Bnot statedArena scoreCC-BY-4.0lmarena
Qwen3.5-397B-A17BOTIS Mock AIME 2024-202582.2%qwen3.5-397b-a17b_nonenoneshare of problems solvedCC-BY-4.0epoch
Qwen3.5-397B-A17BFrontierMath Tiers 1-3 (v2)31.2%qwen3.5-397b-a17b_nonenoneshare of problems solvedCC-BY-4.0epoch
Qwen3.5-397B-A17BGPQA Diamond86.4%qwen3.5-397b-a17b_nonenoneshare of questions correctCC-BY-4.0epoch
Qwen3.5-FlashArena text: business and management1,394Qwen3.5-Flashnot statedArena scoreCC-BY-4.0lmarena
Qwen3.5-FlashArena text: coding1,412Qwen3.5-Flashnot statedArena scoreCC-BY-4.0lmarena
Qwen3.5-FlashArena text: creative writing1,343Qwen3.5-Flashnot statedArena scoreCC-BY-4.0lmarena
Qwen3.5-FlashArena text: following instructions1,375Qwen3.5-Flashnot statedArena scoreCC-BY-4.0lmarena
Qwen3.5-FlashArena text: maths1,408Qwen3.5-Flashnot statedArena scoreCC-BY-4.0lmarena
Qwen3.5-FlashArena text: overall preference1,397Qwen3.5-Flashnot statedArena scoreCC-BY-4.0lmarena
Qwen3.5-FlashArena WebDev: building web apps1,244Qwen3.5-Flashnot statedArena scoreCC-BY-4.0lmarena
Qwen3.5-Max-PreviewArena text: business and management1,463Qwen3.5-Max-Previewnot statedArena scoreCC-BY-4.0lmarena
Qwen3.5-Max-PreviewArena text: coding1,487Qwen3.5-Max-Previewnot statedArena scoreCC-BY-4.0lmarena
Qwen3.5-Max-PreviewArena text: creative writing1,464Qwen3.5-Max-Previewnot statedArena scoreCC-BY-4.0lmarena
Qwen3.5-Max-PreviewArena text: following instructions1,467Qwen3.5-Max-Previewnot statedArena scoreCC-BY-4.0lmarena
Qwen3.5-Max-PreviewArena text: maths1,474Qwen3.5-Max-Previewnot statedArena scoreCC-BY-4.0lmarena
Qwen3.5-Max-PreviewArena text: overall preference1,470Qwen3.5-Max-Previewnot statedArena scoreCC-BY-4.0lmarena
Qwen3.6-Max-PreviewArena text: business and management1,427Qwen3.6-Max-Previewnot statedArena scoreCC-BY-4.0lmarena
Qwen3.6-Max-PreviewArena text: coding1,471Qwen3.6-Max-Previewnot statedArena scoreCC-BY-4.0lmarena
Qwen3.6-Max-PreviewArena text: creative writing1,435Qwen3.6-Max-Previewnot statedArena scoreCC-BY-4.0lmarena
Qwen3.6-Max-PreviewArena text: following instructions1,439Qwen3.6-Max-Previewnot statedArena scoreCC-BY-4.0lmarena
Qwen3.6-Max-PreviewArena text: maths1,464Qwen3.6-Max-Previewnot statedArena scoreCC-BY-4.0lmarena
Qwen3.6-Max-PreviewArena text: overall preference1,447Qwen3.6-Max-Previewnot statedArena scoreCC-BY-4.0lmarena
Qwen3.6-Max-PreviewArena WebDev: building web apps1,482Qwen3.6-Max-Previewnot statedArena scoreCC-BY-4.0lmarena
Qwen3.6-Max-PreviewOTIS Mock AIME 2024-202591.1%qwen3.6-max-previewnot statedshare of problems solvedCC-BY-4.0epoch
Qwen3.6-Max-PreviewGPQA Diamond87.4%qwen3.6-max-previewnot statedshare of questions correctCC-BY-4.0epoch
Qwen3.6-Max-PreviewSimpleQA Verified52.0%qwen3.6-max-previewnot statedshare of questions correctCC-BY-4.0epoch
Qwen3.6-Max-PreviewSWE-bench Verified (Epoch runs)76.7%qwen3.6-max-previewnot statedshare of issues resolvedCC-BY-4.0epoch
Qwen3.6-PlusArena text: business and management1,430Qwen3.6-Plusnot statedArena scoreCC-BY-4.0lmarena
Qwen3.6-PlusArena text: coding1,467Qwen3.6-Plusnot statedArena scoreCC-BY-4.0lmarena
Qwen3.6-PlusArena text: creative writing1,403Qwen3.6-Plusnot statedArena scoreCC-BY-4.0lmarena
Qwen3.6-PlusArena text: following instructions1,424Qwen3.6-Plusnot statedArena scoreCC-BY-4.0lmarena
Qwen3.6-PlusArena text: maths1,449Qwen3.6-Plusnot statedArena scoreCC-BY-4.0lmarena
Qwen3.6-PlusArena text: overall preference1,436Qwen3.6-Plusnot statedArena scoreCC-BY-4.0lmarena
Qwen3.6-PlusArena WebDev: building web apps1,462Qwen3.6-Plusnot statedArena scoreCC-BY-4.0lmarena
Qwen3.6-PlusOTIS Mock AIME 2024-202593.3%qwen3.6-plusnot statedshare of problems solvedCC-BY-4.0epoch
Qwen3.6-PlusFrontierMath Tiers 1-3 (v2)32.3%qwen3.6-plus_nonenoneshare of problems solvedCC-BY-4.0epoch
Qwen3.6-PlusGPQA Diamond88.4%qwen3.6-plusnot statedshare of questions correctCC-BY-4.0epoch
Qwen3.6-PlusSimpleQA Verified44.1%qwen3.6-plusnot statedshare of questions correctCC-BY-4.0epoch
Qwen3.6-PlusSWE-bench Verified (Epoch runs)57.9%qwen3.6-plusnot statedshare of issues resolvedCC-BY-4.0epoch
Qwen3.7-Max-PreviewArena text: business and management1,446Qwen3.7-Max-Previewnot statedArena scoreCC-BY-4.0lmarena
Qwen3.7-Max-PreviewArena text: coding1,497Qwen3.7-Max-Previewnot statedArena scoreCC-BY-4.0lmarena
Qwen3.7-Max-PreviewArena text: creative writing1,446Qwen3.7-Max-Previewnot statedArena scoreCC-BY-4.0lmarena
Qwen3.7-Max-PreviewArena text: following instructions1,460Qwen3.7-Max-Previewnot statedArena scoreCC-BY-4.0lmarena
Qwen3.7-Max-PreviewArena text: maths1,486Qwen3.7-Max-Previewnot statedArena scoreCC-BY-4.0lmarena
Qwen3.7-Max-PreviewArena text: overall preference1,475Qwen3.7-Max-Previewnot statedArena scoreCC-BY-4.0lmarena
Qwen3.7-PlusArena text: business and management1,435Qwen3.7-Plusnot statedArena scoreCC-BY-4.0lmarena
Qwen3.7-PlusArena text: coding1,473Qwen3.7-Plusnot statedArena scoreCC-BY-4.0lmarena
Qwen3.7-PlusArena text: creative writing1,438Qwen3.7-Plusnot statedArena scoreCC-BY-4.0lmarena
Qwen3.7-PlusArena text: following instructions1,440Qwen3.7-Plusnot statedArena scoreCC-BY-4.0lmarena
Qwen3.7-PlusArena text: maths1,467Qwen3.7-Plusnot statedArena scoreCC-BY-4.0lmarena
Qwen3.7-PlusArena text: overall preference1,455Qwen3.7-Plusnot statedArena scoreCC-BY-4.0lmarena
Qwen3.7-PlusOTIS Mock AIME 2024-202580.0%qwen3.7-plus_nonenoneshare of problems solvedCC-BY-4.0epoch
Qwen3.7-PlusFrontierCode (reported by Cognition)10.2%qwen3.7-plusnone, mini-swe-agentmean score over 5 attemptsCC-BY-4.0epoch
Qwen3.7-PlusFrontierMath Tiers 1-3 (v2)34.4%qwen3.7-plus_nonenoneshare of problems solvedCC-BY-4.0epoch
Qwen3.7-PlusGPQA Diamond81.8%qwen3.7-plus_nonenoneshare of questions correctCC-BY-4.0epoch
Qwen3.8-MaxArena text: business and management1,480Qwen3.8-Maxnot statedArena scoreCC-BY-4.0lmarena
Qwen3.8-MaxArena text: coding1,504Qwen3.8-Maxnot statedArena scoreCC-BY-4.0lmarena
Qwen3.8-MaxArena text: creative writing1,478Qwen3.8-Maxnot statedArena scoreCC-BY-4.0lmarena
Qwen3.8-MaxArena text: following instructions1,480Qwen3.8-Maxnot statedArena scoreCC-BY-4.0lmarena
Qwen3.8-MaxArena text: maths1,506Qwen3.8-Maxnot statedArena scoreCC-BY-4.0lmarena
Qwen3.8-MaxArena text: overall preference1,483Qwen3.8-Maxnot statedArena scoreCC-BY-4.0lmarena
Qwen3.8-MaxArena WebDev: building web apps1,672Qwen3.8-Maxnot statedArena scoreCC-BY-4.0lmarena
Qwen3.8-MaxOTIS Mock AIME 2024-202599.4%qwen3.8-max_xhighxhighshare of problems solvedCC-BY-4.0epoch
Qwen3.8-MaxFrontierMath Tiers 1-3 (v2)74.7%qwen3.8-max_xhighxhighshare of problems solvedCC-BY-4.0epoch
Qwen3.8-MaxGPQA Diamond92.7%qwen3.8-max_xhighxhighshare of questions correctCC-BY-4.0epoch
Qwen3.8-MaxSimpleQA Verified45.8%qwen3.8-max_xhighxhighshare of questions correctCC-BY-4.0epoch