whatmodel.sh / Grok 3 Mini Beta
Grok 3 Mini Beta
↗ share
+ contribute
Grok 3 Mini Beta is a closed-weights language model from SpaceXAI (xAI), released April 2025. Licence: proprietary.
Catalogue identity notes
Comparison model from the Epoch AI internal-evaluation CSVs (Epoch release date 2025-04-09). API model grok-3-mini-beta; reasoning effort recorded per result. Not listed on xAI's current models page (observed 2026-09-28).
$ how it compares
The LLMs ranked either side of Grok 3 Mini Beta, out of 828 with a score, then more from SpaceXAI (xAI).
| # | model | score | where | ||
|---|---|---|---|---|---|
| 183 |  Qwen3-VL 235B-A22B Instruct | 64±1 | local | ||
| 184 |  Nova 2 Pro (Preview) | 64±2 | api $3.78/M | ||
| 185 | Grok 3 Mini Beta | 64±1 | comparison only | ||
| 186 |  MiniCPM5 2B | 64±1 | local | ||
| 187 | Grok 3 Beta | 63±1 | comparison only | ||
| more from SpaceXAI (xAI) | |||||
| 30 | Grok 4.6 | 77±1 | api $3.00/M | ||
| 42 | Grok 4.5 | 75±1 | api $3.00/M | ||
| 47 | Grok 4.7 | 75±1 | api $3.00/M | ||
$ what it's good at
agent tasks▁▁▂▆█▆▅▄▄▃▂▁ #270 of 340 * · 9.3% of its score · 2 tests
Beats o3-mini here, though they rank higher overall. Trails Granite 4.0 1B, SmolLM3 3B, Phi-3.5-mini-instruct here, though they rank lower overall.
| test | result | vs the others in its run | counts |
|---|---|---|---|
| Finance Agent Benchmark v1 (537 questions, class-balanced accuracy) | 30.9% ✎ | █·▃▃·█▅▅▅▅▃▅ beats 15/18 | 7.4% |
| BFCL v3 | 53.0% ✎ | ▁▁▁▁▁▃▅█▅▄▁▁ beats 51/114 | 1.9% |
coding▁▁▁▃▅▇█▇▅▄▃▁ #158 of 524 * · 36% of its score · 2 tests
Beats Gemini 2.5 Pro, Gemini 3.1 Flash-Lite, Qwen3-Coder-Next (80B-A3B) here, though they rank higher overall. Trails Devstral Small 2 24B, Llama 3.1 Nemotron Ultra 253B v1, Nemotron 3.5 Lightning 30B-A3B here, though they rank lower overall.
| test | result | vs the others in its run | counts |
|---|---|---|---|
| Aider polyglot (225 exercises, two attempts) | 49.3% ✎ | ▂·▂▃▅▂▃█▆▄▂▂ beats 26/49 | 28% |
| LiveCodeBench official (Aug 1, 2024–May 1, 2025) | 66.7% ✎ | ▃▃▅▃··▃▅█▅▅▃ beats 10/15 | 7.6% |
chat & writing▁▁▁▃▅▄▆█▇▇▇▃ #168 of 463 * · 10% of its score · 1 test
Beats INTELLECT-3, Nemotron 3 Super 120B-A12B, MiniMax-M2 here, though they rank higher overall. Trails LFM2.5 1.2B Thinking, LFM2.5 8B-A1B, LFM2.5 2.6B here, though they rank lower overall.
| test | result | vs the others in its run | counts |
|---|---|---|---|
| LMArena (text) | 1363.5 ✎ | ▁▁▁▂▃▃▃▆▆▇█▃ beats 205/379 | 10% |
math▁▂▃▅▇▅▅█▄▁▁▁ #178 of 462 · 22% of its score · 3 tests
Beats MiniMax-M2.7, MiniMax-M2.1, Mistral Medium 3.5 (128B) here, though they rank higher overall. Trails Olmo 3 7B Think, Nemotron Nano 9B v2, Qwen3-VL 4B Thinking here, though they rank lower overall.
| test | result | vs the others in its run | counts |
|---|---|---|---|
| MATH level 5 | 88.1% ✎ | ▂▂▅▆▄█▆▃▅▂▃▄ beats 75/92 | 11% |
| OTIS mock AIME 2024–2025 | 77.8% ✎ | ▃▃▄▃▃▃▇▇█▃▂▅ beats 111/193 | 9.4% |
| THINK-Bench final-answer accuracy (reasoning models) | 91.8% ✎ | ▄█·▄····▄▄·█ beats 5/7 | 1.4% |
web search▂·▂▇▅█▃▄▁▂▃▁ #4 of 92 * · 7.4% of its score · 1 test
Beats GPT-6 Astra, Claude Opus 5.5, Claude Fable 5.1 here, though they rank higher overall.
| test | result | vs the others in its run | counts |
|---|---|---|---|
| Finance Agent Benchmark v1 (537 questions, class-balanced accuracy) | 30.9% ✎ | █·▃▃·█▅▅▅▅▃▅ beats 15/18 | 7.4% |
other tests2 not in a task · 23% of its score
| test | result | vs the others in its run | counts |
|---|---|---|---|
| GPQA Diamond | 75.5% ✎ | ▁▁▁▁▂▄▆▄▅▇█▅ beats 113/208 | 17% |
| EQ-Bench 3 (emotional intelligence, judge Elo) | 1146.8 ✎ | ▁·▂▁▄▅▅▃█▄▁▁ beats 19/62 | 5.9% |
$ where to run it
No verified API or download recorded yet · add one
$ how the score was worked out · Epoch AI — internal evaluatio… 37%, Aider polyglot benchmark runs 28%, LMArena leaderboard dataset 10% and 5 more
Each result implies a score on its own, after adjusting for how hard its test is and who ran it. The score is their weighted average, with a small pull toward the middle for thinly tested models. Each test's weight is its counts column above; here it's added up by who ran the tests. How it works
| source | results | share of the score |
|---|---|---|
 Epoch AI — internal evaluations | 3 | 37% |
 Aider polyglot benchmark runs | 1 | 28% |
 LMArena leaderboard dataset | 1 | 10% |
| LiveCodeBench official website results (2025-08-01) | 1 | 8% |
 Finance Agent Benchmark (Vals AI, arXiv:2508.00828v1) | 1 | 7% |
 EQ-Bench 3 canonical leaderboard results | 1 | 6% |
| Berkeley Function Calling Leaderboard result archive | 1 | 2% |
| THINK-Bench: Evaluating Thinking Efficiency and Chain-of-Thought Quality of Large Reasoning Models (arXiv:2505.22113v1) | 1 | 1% |
| prior · pull toward the middle | 0% | |
| weighted average | = 63.5 |
$ improve this page
Use a short GitHub form or edit the data through a pull request. Every number needs a source; here's which ones we can use.
data updated · sources and licences · json