Updated daily at 05:00 UTC
- Released
- 9 Jul 2025
- Status
- Retired
- Retired
- 15 May 2026
- Context
- 256K
- Max output
- 64K
- Input per 1M tokens
- $3.00
- Output per 1M tokens
- $15.00
- Open weights
- No
- License
- proprietary
- Parameters
- —
- Knowledge cutoff
- —
- Reasoning
- Yes
- Modalities
- text, image → text
Category scores
Out of 100, from independent and third-party results only, by method category-2026-09-29.1.
| Category | Score | Rank | Coverage |
|---|---|---|---|
| Knowledge & reasoning | 38.4 | 38 | 2 of 4 |
| Coding & software engineering | 79.6 | Not ranked | 1 of 5 |
| Agentic tool use | 45.5 | 5 | 2 of 2 |
| Human preference | 51.6 | Not ranked | 1 of 5 |
Benchmark results
Every result we hold for this model, with who ran it and where it was published.
Knowledge & reasoning
| Benchmark | Metric | Value | Provenance | Setting | Evaluated | Source |
|---|---|---|---|---|---|---|
| Chess Puzzles | Puzzles solved | 28.0% | Independent | — | 30 Jan 2026 | Epoch AI |
| SimpleBench | Score (AVG@5) | 60.5% | Third party | — | — | https://lmcouncil.ai/benchmarks via Epoch AI |
Scientific reasoning
| Benchmark | Metric | Value | Provenance | Setting | Evaluated | Source |
|---|---|---|---|---|---|---|
| GPQA Diamond | Accuracy | 87.0% | Independent | — | — | Epoch AI |
Mathematics
| Benchmark | Metric | Value | Provenance | Setting | Evaluated | Source |
|---|---|---|---|---|---|---|
| FrontierMath Tier 4 (v1) | Accuracy | 2.1% | Independent | — | 11 Aug 2025 | Epoch AI |
| FrontierMath Tiers 1-3 (v1) | Accuracy | 19.7% | Independent | — | 13 Nov 2025 | Epoch AI |
| OTIS Mock AIME 2024-2025 | Accuracy | 84.0% | Independent | — | — | Epoch AI |
Coding & software engineering
| Benchmark | Metric | Value | Provenance | Setting | Evaluated | Source |
|---|---|---|---|---|---|---|
| Aider Polyglot | Percent correct | 79.6% | Third party | — | — | Aider LLM Leaderboards via Epoch AI |
| Aider Polyglot | Percent correct | 79.6% | Third party | high | — | https://aider.chat/docs/leaderboards/ via Epoch AI |
| Cybench | Unguided challenges solved | 43.0% | Self-reported | — | — | Grok 4 Model Card via Epoch AI |
| METR Time Horizons | Average task success (Epoch's column) | 66.6% | Third party | — | — | METR - Measuring AI Ability to Complete Long Tasks via Epoch AI |
| Terminal-Bench 2.0 | Accuracy (mean over trials) | 27.2% | Third party | — | — | Terminal-Bench v2 Leaderboard via Epoch AI |
| Terminal-Bench 2.0 | Accuracy (mean over trials) | 27.2% | Third party | — | — | https://www.tbench.ai/leaderboard/terminal-bench/2.0 via Epoch AI |
| Terminal-Bench 2.0 | Accuracy (mean over trials) | 25.4% | Third party | — | — | Terminal-Bench v2 Leaderboard via Epoch AI |
| Terminal-Bench 2.0 | Accuracy (mean over trials) | 25.4% | Third party | — | — | https://www.tbench.ai/leaderboard/terminal-bench/2.0 via Epoch AI |
| Terminal-Bench 2.0 | Accuracy (mean over trials) | 23.1% | Third party | — | — | https://www.tbench.ai/leaderboard/terminal-bench/2.0 via Epoch AI |
| Terminal-Bench 2.0 | Accuracy (mean over trials) | 23.1% | Third party | — | — | Terminal-Bench v2 Leaderboard via Epoch AI |
| WeirdML (v2) | Average accuracy | 45.7% | Third party | — | — | https://htihle.github.io/weirdml.html via Epoch AI |
Agentic tool use
| Benchmark | Metric | Value | Provenance | Setting | Evaluated | Source |
|---|---|---|---|---|---|---|
| BALROG | Average progress | 43.6% | Third party | — | — | https://balrogai.com/ via Epoch AI |
| Deep Research Bench (FutureSearch) | Average score | 47.3% | Third party | — | — | https://evals.futuresearch.ai/#drb via Epoch AI |
Long context
| Benchmark | Metric | Value | Provenance | Setting | Evaluated | Source |
|---|---|---|---|---|---|---|
| Fiction.LiveBench | % correct at 16K tokens (Epoch's column) | 94.4% | Third party | — | — | Fiction.live leaderboard via Epoch AI |
Vision & multimodal understanding
| Benchmark | Metric | Value | Provenance | Setting | Evaluated | Source |
|---|---|---|---|---|---|---|
| GeoBench | Country accuracy | 45.0% | Third party | — | — | GeoBench leaderboard via Epoch AI |
Writing
| Benchmark | Metric | Value | Provenance | Setting | Evaluated | Source |
|---|---|---|---|---|---|---|
| LLM Creative Story-Writing Benchmark V1 (Lech Mazur) | Mean rubric score (percent of the 10-point scale) | 76.9% | Third party | — | — | lechmazur/writing Github repository via Epoch AI |
Human preference
| Benchmark | Metric | Value | Provenance | Setting | Evaluated | Source |
|---|---|---|---|---|---|---|
| Text Arena (arena.ai) | Arena score (Elo scale) | 1411 | Independent | — | 25 Sep 2026 | Arena |
| Text Arena (arena.ai) | Votes | 39389 | Independent | — | 25 Sep 2026 | Arena |
| Vision Arena (arena.ai) | Arena score (Elo scale) | 1184 | Independent | — | 28 Sep 2026 | Arena |
| Vision Arena (arena.ai) | Votes | 30195 | Independent | — | 28 Sep 2026 | Arena |
Other Grok versions
- Grok 4.721 Sep 2026
- Grok 4.612 Aug 2026
- Grok 4.58 Jul 2026
- Grok 4.317 Apr 2026
- Grok 4.20 Multi-Agent9 Mar 2026
- Grok 4.20Feb 2026
- Grok 4.1 Fast19 Nov 2025
- Grok 4.117 Nov 2025
- Grok 4 Fast19 Sep 2025
- Grok 317 Feb 2025
- Grok 3 Mini17 Feb 2025
- Grok 2 Vision12 Dec 2024
- Grok BetaNov 2024
- Grok Vision BetaNov 2024
- Grok 214 Aug 2024
- Grok 2 mini14 Aug 2024
- Grok-1.515 May 2024
- Grok-13 Nov 2023