xAIReasoning modelGrok 4 Fast
Best configuration ranks #234 of 610 on the BenchLeader Index at 53.0 ±4.2 (thinking reasoning effort). Last measured 8 Sept 2026. Released 19 Sept 2025.
- Blended price
- $1.56/M
- $1.25 in · $2.50 out
- Output speed
- –
- First answer
- –
- Context
- 2M
- Overall index53
- Reasoning59
- Coding42
- Agents & tools50
- Maths58
- Knowledge54
- Instruction following52
- Human preference59
- Multimodal46
- Long context63
- Composite52
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. “Default” means the publisher did not say which setting was used.
| Effort | Index | Rank | Speed | First answer | Chat reply cost | Categories |
|---|---|---|---|---|---|---|
| no reasoning | 41.9 | #473 | – | – | $0.0013 | Coding 32 · Knowledge 43 · Maths 42 |
| thinkingbest | 53.0 | #234 | – | – | $0.0013 | Agents & tools 50 · Coding 42 · Composite 52 · Human preference 59 · Instruction following 52 · Knowledge 54 · Long context 63 · Maths 58 · Multimodal 46 · Reasoning 59 |
| default | 46.5 | #368 | – | – | $0.0002 | Agents & tools 47 · Coding 52 · Composite 43 · Human preference 61 · Instruction following 40 · Knowledge 38 · Long context 50 · Multimodal 32 · Reasoning 47 |
Benchmark results
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 | no reasoning | thinking | default | Source | Trend |
|---|---|---|---|---|---|
| LMArena Hard Prompts | – | 1414#142 | 1431#119 | LMArena | |
| GPQA Diamond (AA)not in index | – | 84.8%#132 | 60.6%#352 | Artificial Analysis | |
| Humanity's Last Exam (AA)not in index | – | 19.1%#167 | 4.5%#414 | Artificial Analysis | |
| GPQA Diamond (Vals)not in index | 62.1%#105 | 85.3%#47 | – | Vals AI | |
| Kagi LLM Benchmark | – | 66.1%#38 | 35.4%#116 | Kagi LLM Benchmark |
Coding
| Benchmark | no reasoning | thinking | default | Source | Trend |
|---|---|---|---|---|---|
| WeirdML | – | – | 42.9%#89 | WeirdML | |
| LMArena Coding | – | 1438#141 | 1458#117 | LMArena | |
| LMArena WebDev | – | 1162#117 | – | LMArena | |
| LiveCodeBench | 46.1%#115 | 79.0%#73 | – | Vals AI | |
| IOI | 3.8%#49 | 11.5%#32 | – | Vals AI | |
| SWE-bench (Vals)not in index | – | 45.4%#80 | – | Vals AI |
Agents & tools
| Benchmark | no reasoning | thinking | default | Source | Trend |
|---|---|---|---|---|---|
| Cybench | – | – | 30.0%#9 | Cybench | |
| Terminal-Bench Hard | – | 18.9%#161 | 12.1%#202 | Artificial Analysis | |
| τ²-Bench Telecom (AA)not in index | – | 65.8%#161 | 63.7%#167 | Artificial Analysis |
Maths
| Benchmark | no reasoning | thinking | default | Source | Trend |
|---|---|---|---|---|---|
| AIME (Vals) | 33.3%#70 | 91.3%#25 | – | Vals AI | |
| MGSM | 88.0%#52 | 90.9%#32 | – | Vals AI |
Knowledge
| Benchmark | no reasoning | thinking | default | Source | Trend |
|---|---|---|---|---|---|
| AA-Omniscience | – | -29.9#216 | -54.4#354 | Artificial Analysis | |
| MMLU-Pro | 70.3%#113 | 79.7%#90 | – | Vals AI | |
| LegalBench | 78.4%#93 | 80.6%#77 | – | Vals AI | |
| CorpFin | 58.4%#81 | 66.9%#15 | – | Vals AI | |
| TaxEval | 71.6%#75 | 75.7%#13 | – | Vals AI | |
| MedQA | 75.4%#76 | 92.1%#35 | – | Vals AI |
Instruction following
| Benchmark | no reasoning | thinking | default | Source | Trend |
|---|---|---|---|---|---|
| IFBench | – | 50.5%#169 | 37.7%#276 | Artificial Analysis |
Human preference
| Benchmark | no reasoning | thinking | default | Source | Trend |
|---|---|---|---|---|---|
| LMArena Text | – | 1405#126 | 1418#107 | LMArena |
Multimodal
| Benchmark | no reasoning | thinking | default | Source | Trend |
|---|---|---|---|---|---|
| MMMU-Pro | – | 61.8%#171 | 48.1%#212 | Artificial Analysis |
Long context
| Benchmark | no reasoning | thinking | default | Source | Trend |
|---|---|---|---|---|---|
| Fiction.LiveBench 120k | – | – | 75.0%#7 | Fiction.live | |
| AA-LCR | – | 73.7%#125 | 24.0%#344 | Artificial Analysis |
Composite
| Benchmark | no reasoning | thinking | default | Source | Trend |
|---|---|---|---|---|---|
| Epoch Capabilities Indexnot in index | – | – | 144.2#75 | Epoch AI Benchmarking Hub | |
| AA Intelligence Index | – | 17.9#194 | 11.1#296 | Artificial Analysis |
What a task costs
Estimates from list price, output speed and time to first answer for the best configuration. “With caching” assumes three-quarters of the input is served from the prompt cache. Reasoning tokens are not modelled.
| Workload | Tokens in / out | Cost | With caching | Time |
|---|---|---|---|---|
| Chat reply | 400 / 300 | $0.0013 | – | – |
| Summarise a 30-page report | 12,000 / 600 | $0.017 | – | – |
| Code edit | 6,000 / 1,500 | $0.011 | – | – |
| Agentic coding session | 60,000 / 4,000 | $0.085 | – | – |
| Structured extraction | 2,000 / 200 | $0.0030 | – | – |
See also
Data as of 9 Sept 2026. Compare these configurations.