SpaceXAIGrok 3 mini
Grok 3 mini is a SpaceXAI proprietary reasoning model, released 9 Apr 2025. Its best configuration (high reasoning effort) ranks #310 of 785 on the BenchLeader Index at 52.6 ±3.7, in the upper half. The ± is the point: 287 other configurations score within that range, so they and this one cannot be told apart on quality alone — price and speed are what separate them. It scores highest in maths (59) and lowest in coding (48). It has been measured at 5 reasoning-effort settings; this summary describes the best-scoring one, and the tabs above switch between them. Last measured 22 Oct 2026.
- Blended price
- –
- Output speed
- –
- First answer
- –
- Context
- –
- Full answer
- –
- Cost per run
- –
- Released
- 9 Apr 2025
- Overall index53
- Reasoning54
- Coding48
- Maths59
- Human preference53
Versions
SpaceXAI has shipped 2 models under this name. Each is ranked on its own results; a newer version often has fewer results so far, which holds its index nearer the average until more arrive.
| Model | Released | Index | Rank |
|---|---|---|---|
| Grok 3 minihighthis page | 9 Apr 2025 | 52.6 | #310 |
| Grok 2 Mini | – | – | – |
Reasoning-effort configurations
The same model behaves differently depending on how much it is allowed to think. Each row is one setting, scored only on the benchmarks that were run at that setting. “Not stated” collects results from publishers that did not say which setting they used; for a reasoning model that is usually its thinking mode, but we do not assume it. Pick a setting here or at the top of the page to see its price, speed and category scores.
| Effort | Index | Rank | Speed | First answer | Chat reply cost | Categories |
|---|---|---|---|---|---|---|
| low | 49.8 | #387 | – | – | – | Maths 59 · Reasoning 41 |
| medium | – | – | – | – | – | Long context 56 |
| highbest | 52.6 | #310 | – | – | – | Coding 48 · Human preference 53 · Maths 59 · Reasoning 54 |
| thinking | 49.4 | #399 | 68 tok/s | 31 s | $0.0003 | Agents & tools 49 · Composite 46 · Instruction following 48 · Knowledge 58 · Long context 52 · Reasoning 42 |
| not stated | 52.3 | #321 | – | – | – | Coding 51 · Human preference 52 · Instruction following 58 · Maths 49 · Reasoning 54 |
How its index has moved
8 Sept 2026 to 23 Oct 2026- high
- not stated
- low
- thinking
The index is recomputed from scratch every day, so a line moves when a new benchmark result lands, when a publisher revises a score, or when the models it is normalised against change. Early movement usually means the score is still settling.
Benchmark results
See all LLM benchmarks. One column per reasoning effort. Rank is among every configuration of every model on that benchmark. Hover a score for the run it came from.
Reasoning
| Benchmark | low | high | thinking | not stated | Source | Trend |
|---|---|---|---|---|---|---|
| GPQA Diamond | 76.3%#142 | 75.5%#149 | – | – | Epoch AI Benchmarking Hub | |
| Arena Hard Prompts | – | 1378#189 | – | 1370#197 | Arena (LMArena) | |
| GPQA Diamond (AA)not in index | – | – | 79.1%#202 | – | Artificial Analysis | |
| Humanity's Last Exam (AA)not in index | – | – | 11.0%#303 | – | Artificial Analysis | |
| Kagi LLM Benchmarkeffort not stated | – | – | – | 61.3%#47 | Kagi LLM Benchmark | |
| GPQA Diamond (HELM)not in indexeffort not stated | – | – | – | 67.5%#14 | HELM Capabilities | |
| ARC-AGI-1 | 16.5%#226 | – | – | – | ARC Prize | |
| ARC-AGI-2 | 0.4%#239 | – | – | – | ARC Prize | |
| CritPt | – | – | 0.6%#254 | – | Artificial Analysis |
Coding
| Benchmark | high | not stated | Source | Trend |
|---|---|---|---|---|
| WeirdML | 42.6%#93 | – | WeirdML | |
| Arena Coding | 1393#199 | 1386#207 | Arena (LMArena) | |
| Aider Polygloteffort not stated | – | 49.3%#26 | Aider polyglot leaderboard |
Agents & tools
| Benchmark | thinking | not stated | Source | Trend |
|---|---|---|---|---|
| Terminal-Bench Hard | 17.4%#178 | – | Artificial Analysis | |
| τ²-Bench Telecom (AA)not in index | 90.3%#64 | – | Artificial Analysis | |
| BrowseCompnot in indexeffort not stated | – | 1.7%#10 | OpenAI |
Maths
| Benchmark | low | high | not stated | Source | Trend |
|---|---|---|---|---|---|
| OTIS Mock AIME | 62.2%#159 | 77.8%#122 | – | Epoch AI Benchmarking Hub | |
| MATH Level 5 | 90.9%#20 | 88.1%#24 | – | Epoch AI Benchmarking Hub | |
| Omni-MATH (HELM)effort not stated | – | – | 31.8%#41 | HELM Capabilities | |
| Arena Maths | – | 1389#164 | 1370#182 | Arena (LMArena) |
Knowledge
| Benchmark | thinking | not stated | Source | Trend |
|---|---|---|---|---|
| AA-Omniscience | -6.8#171 | – | Artificial Analysis | |
| MMLU-Pro (HELM)not in indexeffort not stated | – | 79.9%#19 | HELM Capabilities | |
| AA-Omniscience: accuracynot in index | 15.1%#420 | – | Artificial Analysis | |
| AA-Omniscience: non-hallucinationnot in index | 74.2%#28 | – | Artificial Analysis |
Instruction following
| Benchmark | high | thinking | not stated | Source | Trend |
|---|---|---|---|---|---|
| IFBench | – | 45.9%#198 | – | Artificial Analysis | |
| IFEval (HELM)effort not stated | – | – | 95.1%#1 | HELM Capabilities | |
| WildBench (HELM)effort not stated | – | – | 65.1%#65 | HELM Capabilities | |
| Arena Instruction Followingnot in index | 1361#181 | – | 1353#187 | Arena (LMArena) |
Human preference
| Benchmark | low | high | not stated | Source | Trend |
|---|---|---|---|---|---|
| Arena Text | – | 1364#188 | 1358#193 | Arena (LMArena) | |
| Arena Creative Writingnot in index | – | 1332#187 | 1336#181 | Arena (LMArena) | |
| Arena Multi-turnnot in index | – | 1349#195 | 1345#200 | Arena (LMArena) | |
| Arena Longer Queriesnot in index | – | 1377#181 | 1367#190 | Arena (LMArena) | |
| Lech Mazur Creative Writingnot in index | 73.5%#28 | – | – | Lech Mazur |
Long context
| Benchmark | medium | thinking | Source | Trend |
|---|---|---|---|---|
| Fiction.LiveBench 120k | 65.6%#10 | – | Fiction.live | |
| AA-LCR | – | 55.3%#298 | Artificial Analysis |
Composite
| Benchmark | thinking | not stated | Source | Trend |
|---|---|---|---|---|
| Epoch Capabilities Indexnot in indexeffort not stated | – | 140.2#110 | Epoch AI Benchmarking Hub | |
| AA Intelligence Index v4.3.2 | 14.6#293 | – | Artificial Analysis | |
| HELM Capabilities meannot in indexeffort not stated | – | 67.9%#26 | HELM Capabilities |
More from SpaceXAI
See also
Questions people ask
- What are Grok 3 mini's benchmark scores?
- As of 25 Oct 2026, Grok 3 mini (high effort) scores 52.6 ±3.7 on the BenchLeader Index, strongest in maths (59) and weakest in coding (48). That rests on 8 independent benchmark results, and every score on this page links to its publisher.
- How does Grok 3 mini rank?
- Grok 3 mini (high effort) is #310 of 785 ranked configurations on 25 Oct 2026. 287 other configurations score within its margin, so the exact position matters less than the band. The leader is Claude Opus 5.5 (max effort) at 72.6.
- When was Grok 3 mini released?
- SpaceXAI released Grok 3 mini on 9 Apr 2025. BenchLeader has results for it at 5 reasoning-effort settings.
Data as of 25 Oct 2026. Compare these configurations.
Cite as: BenchLeader, “Grok 3 mini: benchmarks, pricing, speed and rank”, data as of 25 Oct 2026, https://www.benchleader.com/models/grok-3-mini.