Grok 4.20 Beta 1
Made by xAI. Epoch AI does not state how it can be accessed. Appears solid at writing among the models listed here.
In short
What it does well
- Writing: Strong (39th of 77 listed models)
Where it falls short
- Coding: Behind the leaders (72nd of 77 listed models)
What we don't know yet
- Web-building votes: Not enough data
- Hard maths problems: Not enough data
- School-competition maths: Not enough data
- Graduate-level science questions: Not enough data
- Short factual questions: Not enough data
- Fixing real software bugs: Not enough data
- Difficult coding tasks: Not enough data
The four ratings
- Top tier
- Strong
- Middle
- Behind the leaders
- Not enough data
Ratings compare only the 77 models listed here, in quarters, best first. “Not enough data” means no result yet. It is never a low score.
- Writing
- StrongHow often people preferred its answers to writing prompts. 1,438 on Arena's scale · Source for Grok 4.20 Beta 1 writing score (lmarena)
- Coding
- Behind the leadersHow often people preferred its answers to coding questions. 1,447 on Arena's scale · Source for Grok 4.20 Beta 1 coding score (lmarena)
- Maths
- Not enough dataNo result yet for hard maths problems.
- Knowledge
- Not enough dataNo result yet for short factual questions.
Evidence completeness: Thin evidence, 6 of 13 tests have a result.
Where it is relatively stronger or weaker
This compares each kind of request with the model's own overall position among the 77 listed models, using Arena's votes. It is about performance on different languages, kinds of request and topics, not about political or social bias.
Relatively stronger
- German: 27th of 55 on questions asked in German, against 58th of 77 overall. Source for German
Relatively weaker
No category stands out as clearly weaker than its overall position.
Show every category we can see
| Category | Position | Arena score | Votes | Source |
|---|---|---|---|---|
| Expert-level prompts Kind of request | 72nd of 77 | 1,427 95% range 1,415 to 1,439 | 2,472 | Source for Expert-level prompts |
| Hard prompts Kind of request | 69th of 77 | 1,441 95% range 1,435 to 1,446 | 17,448 | Source for Hard prompts |
| Long prompts Kind of request | 68th of 77 | 1,427 95% range 1,420 to 1,433 | 10,449 | Source for Long prompts |
| Longer conversations Kind of request | 52nd of 77 | 1,451 95% range 1,442 to 1,460 | 4,596 | Source for Longer conversations |
| Chinese Language | 62nd of 76 | 1,476 95% range 1,461 to 1,490 | 1,744 | Source for Chinese |
| English prompts Language | 58th of 77 | 1,451 95% range 1,445 to 1,457 | 12,790 | Source for English prompts |
| French Language | 55th of 72 | 1,455 95% range 1,434 to 1,477 | 859 | Source for French |
| German Language | 27th of 55 | 1,464 95% range 1,436 to 1,491 | 443 | Source for German |
| Korean Language | 33rd of 54 | 1,416 95% range 1,388 to 1,444 | 469 | Source for Korean |
| Non-English prompts Language | 52nd of 77 | 1,433 95% range 1,427 to 1,438 | 15,099 | Source for Non-English prompts |
| Russian Language | 46th of 77 | 1,449 95% range 1,439 to 1,460 | 3,302 | Source for Russian |
| Spanish Language | 57th of 70 | 1,438 95% range 1,419 to 1,458 | 952 | Source for Spanish |
| Entertainment, sport and media Topic | 54th of 77 | 1,417 95% range 1,409 to 1,426 | 5,377 | Source for Entertainment, sport and media |
| Law and government Topic | 65th of 76 | 1,433 95% range 1,420 to 1,446 | 2,107 | Source for Law and government |
| Medicine and health Topic | 50th of 75 | 1,452 95% range 1,438 to 1,465 | 2,015 | Source for Medicine and health |
| Science Topic | 66th of 77 | 1,441 95% range 1,432 to 1,450 | 4,610 | Source for Science |
| Software and IT Topic | 68th of 77 | 1,452 95% range 1,446 to 1,459 | 10,827 | Source for Software and IT |
Bias: what we can and can't see
Our sources do not measure political lean, stereotyping, flattery (agreeing with you too readily) or safety behaviour, so we cannot say how any model does on those. What we can show is above: where a model is relatively stronger or weaker across languages, kinds of request and topics. Also, people voting in Arena can favour answer style, such as length and formatting, and our numbers do not adjust for that. See what AI bias means and what these tests can tell you.
Best-supported use cases
- Coding and building things: not ranked, because there is not enough data for the main test.
- Homework and maths: not ranked, because there is not enough data for the main test.
- Research and checking facts: not ranked, because there is not enough data for the main test.
Where an alternative might be more suitable
- Writing and creating content: rated Strong on writing votes. Compare it with Claude Opus 5.5, the best-supported choice for this task.
- Coding and building things: rated Behind the leaders on coding votes. Compare it with Claude Opus 5.5, the best-supported choice for this task.
- Homework and maths: no result on hard maths problems, so we cannot say. See which models the evidence does support.
- Business emails and everyday work: rated Behind the leaders on business-task votes. Compare it with Gemini 4 Argon, the best-supported choice for this task.
- Research and checking facts: no result on short factual questions, so we cannot say. See which models the evidence does support.
What it is
- Maker
- xAI
- Access, as listed by Epoch AI
- Not stated
- Licence, as listed by Arena
- Proprietary
- Last observation in our data
- 28 Oct 2026
- Data last changed
- 29 Oct 2026, 14:36 UTC
We have not verified an official product page for this model, so we do not link to one.
How you can use it
Proprietary. The model's files are not published, so it is generally available only through its maker or partners the maker chooses.
How you can reach it (an app, a website, a developer API or a third-party tool) and what that costs are not covered by our benchmark data. Where we could verify prices from the maker's own pages, they are shown under Price. See models, apps and harnesses and how AI pricing works.
Price
Price not independently verified
See Grok (checked 29 Oct 2026). Page opened and read. It says Grok is 'free to try on the web and in the apps' and that SuperGrok adds higher limits. It states no prices or model names. The separate plans page opens as a sign-up pop-up showing introductory offers and no model names, so we record nothing else.
We never infer subscription prices from API prices. New to “tokens” and “per million”? See how AI pricing works.
Good to remember
- Benchmarks are samples of tasks. None proves performance in every situation.
- Arena scores come from people choosing between answers. That measures preference, not correctness.
- Ratings are relative to the 77 models listed here. A different set would change them.
- Sources test models at different settings. Each result shows the setting used, and we do not adjust it.
The evidence, for the curious
Show all 13 tests with settings and sources
| Test | Result | Rating | Setting used | Source identifier | Observed | Links |
|---|---|---|---|---|---|---|
| Arena text: overall preference Arena 'text' leaderboard (Bradley-Terry rating, no style control) |
1,444 Arena score 95% range 1,440 to 1,449, 27,889 votes |
Middle | effort: not stated, bradley-terry rating | Grok 4.20 Beta 1 lmarena · text/latest |
28 Oct 2026 | Source for Grok 4.20 Beta 1 Arena text: overall preference score (lmarena) |
| Arena text: creative writing Arena 'text' leaderboard, category creative_writing |
1,438 Arena score 95% range 1,428 to 1,448, 4,171 votes |
Strong | effort: not stated, bradley-terry rating | Grok 4.20 Beta 1 lmarena · text/latest |
28 Oct 2026 | Source for Grok 4.20 Beta 1 Arena text: creative writing score (lmarena) |
| Arena text: coding Arena 'text' leaderboard, category coding |
1,447 Arena score 95% range 1,439 to 1,454, 7,483 votes |
Behind the leaders | effort: not stated, bradley-terry rating | Grok 4.20 Beta 1 lmarena · text/latest |
28 Oct 2026 | Source for Grok 4.20 Beta 1 Arena text: coding score (lmarena) |
| Arena text: maths Arena 'text' leaderboard, category math |
1,432 Arena score 95% range 1,418 to 1,447, 1,683 votes |
Behind the leaders | effort: not stated, bradley-terry rating | Grok 4.20 Beta 1 lmarena · text/latest |
28 Oct 2026 | Source for Grok 4.20 Beta 1 Arena text: maths score (lmarena) |
| Arena text: following instructions Arena 'text' leaderboard, category instruction_following |
1,415 Arena score 95% range 1,408 to 1,422, 8,839 votes |
Behind the leaders | effort: not stated, bradley-terry rating | Grok 4.20 Beta 1 lmarena · text/latest |
28 Oct 2026 | Source for Grok 4.20 Beta 1 Arena text: following instructions score (lmarena) |
| Arena text: business and management Arena 'text' leaderboard, category industry_business_and_management_and_financial_operations |
1,425 Arena score 95% range 1,416 to 1,433, 5,357 votes |
Behind the leaders | effort: not stated, bradley-terry rating | Grok 4.20 Beta 1 lmarena · text/latest |
28 Oct 2026 | Source for Grok 4.20 Beta 1 Arena text: business and management score (lmarena) |
| Arena WebDev: building web apps | Not enough data | |||||
| FrontierMath Tiers 1-3 (v2) | Not enough data | |||||
| OTIS Mock AIME 2024-2025 | Not enough data | |||||
| GPQA Diamond | Not enough data | |||||
| SimpleQA Verified | Not enough data | |||||
| SWE-bench Verified (Epoch runs) | Not enough data | |||||
| FrontierCode (reported by Cognition) | Not enough data | |||||
Missing evidence
Not enough data for these tests:
- Arena WebDev: building web apps
- FrontierMath Tiers 1-3 (v2)
- OTIS Mock AIME 2024-2025
- GPQA Diamond
- SimpleQA Verified
- SWE-bench Verified (Epoch runs)
- FrontierCode (reported by Cognition)
Missing results are not low scores and do not count against the model.
Where a source gives no public run log, we cannot show the run behind the score. See data sources.
Not sure this is the right one?
Tell us what you need help with and we'll narrow the options down.