MiniMax M3 vs Mistral Small 4
MiniMax M3 leads Mistral Small 4 on the Artificial Analysis Intelligence Index, 29.2 to 11.3, and Mistral Small 4 costs 2.0x less per output token ($0.60 vs $1.20 per million). Updated October 23, 2026; 14 benchmarks have a result for both.
MiniMax M3 or Mistral Small 4: which to pick
Only MiniMax M3 has a coding score, 39.2 on the Vibe Coding Index; Mistral Small 4 has none, so the two cannot be ranked on coding.
Pick Mistral Small 4 for output tokens 2.0x cheaper ($0.60 vs $1.20 per million).
- Shared results
- 14
- MiniMax M3 only
- 103
- Mistral Small 4 only
- 1
- Like-for-like categories
- 5 of 9
Only benchmarks that tested both models count toward a pick; a result only one model has is shown but not compared.
Which one for your work
Each card is decided by one number both models carry. Where only one has it, the card says so and makes no call.
Coding
Vibe Coding Index, six public coding benchmarks
No call
Only MiniMax M3 has a coding score, so no call is made.
One side not measured
Intelligence
AA Intelligence Index
MiniMax M3
MiniMax M3 leads, 29.2 against 11.3.
Same board, both measured
Agentic work
AA Agentic Index
MiniMax M3
MiniMax M3 leads, 29.5 against 0.8.
Same board, both measured
Reasoning
GPQA Diamond
No call
Only MiniMax M3 has a GPQA Diamond result, 92.7%.
Directional only
Knowledge
MMLU Pro
No call
Only MiniMax M3 has a MMLU Pro result, 84.2%.
Directional only
Web design
Design Arena: websites
No call
Only MiniMax M3 has a Design Arena: websites result, 1257.
Directional only
Long documents
Listed context window
MiniMax M3
MiniMax M3 takes 1.05M tokens against 262K.
Listed by OpenRouter
Everyday cost
1K input + 500 output tokens
Mistral Small 4
MiniMax M3 $0.0009, Mistral Small 4 $0.0004 per turn.
List prices
Cost per solved task
One SWE-bench Verified task, solved
No call
Only MiniMax M3 has the number, $0.56.
Directional only
At a glance
- Coding index
- MiniMax M3 39.2 from 6 benchmarks, Mistral Small 4 not benchmarked.
- Widest category
- Agentic work, 38.2 points apart on average over 2 shared benchmarks.
- Widest single gap
- τ²-Bench Telecom: MiniMax M3 88.9% against Mistral Small 4 41.2%, 47.7 points apart on a 0 to 100 scale.
- Output price
- MiniMax M3 $1.20 against Mistral Small 4 $0.60 per million output tokens, 2.0x apart.
- Context window
- MiniMax M3 takes 1.05M tokens, Mistral Small 4 262K.
Benchmark by benchmark
First the score: boards each model leads, per category, where both have a result. Then every board, a category to a fold; bars are each result on a 0 to 100 scale, so compare the two in a row, not one row with another.
| Category | MiniMax M3 | Bars | Mistral Small 4 | Ties |
|---|---|---|---|---|
| Coding | 5 | 0 | ||
| Intelligence | 1 | 0 | ||
| Agentic work | 2 | 0 | ||
| Reasoning | 2 | 0 | ||
| General | 4 | 0 | ||
| All shared boards | 14 | 0 |
CodingLike-for-likeSix of these boards feed the Vibe Coding Index50
| Benchmark | MiniMax M3 | Mistral Small 4 | Leader |
|---|---|---|---|
| LiveBench coding | 70.9% | Not measured | Only MiniMax M3 measured |
| LMArena WebDev | 1482 | Not measured | Only MiniMax M3 measured |
| AA Coding Index | 58.6 | 26.6 | MiniMax M3 |
| LiveCodeBench (Vals AI) | 82.2% | Not measured | Only MiniMax M3 measured |
| SWE-bench Verified (Vals AI) | 75.0% | Not measured | Only MiniMax M3 measured |
| Terminal-Bench 4 (Vals AI) | 1.0% | Not measured | Only MiniMax M3 measured |
| Vibe Code Bench (Vals AI) | 47.6% | 0.0% | MiniMax M3 |
| Code migration (Vals AI) | 19.9% | Not measured | Only MiniMax M3 measured |
| Terminal-Bench 2.1 (Vals AI) | 53.6% | Not measured | Only MiniMax M3 measured |
| SWE-bench Multimodal (Vals AI) | 33.3% | Not measured | Only MiniMax M3 measured |
| Vibe Code Bench 1-100 (Vals AI) | 9.2% | Not measured | Only MiniMax M3 measured |
| SkillsBench (Vals AI) | 51.5% | Not measured | Only MiniMax M3 measured |
| CyberBench (Vals AI) | 65.0% | Not measured | Only MiniMax M3 measured |
| SciCode | 47.1% | 38.8% | MiniMax M3 |
| Terminal-Bench Hard (Artificial Analysis run) | 42.4% | 17.4% | MiniMax M3 |
| FrontierCode | 14.7% | Not measured | Only MiniMax M3 measured |
| ALE-Bench | 640.0 | 497.6 | MiniMax M3 |
| LMArena Text: coding | 1470 | Not measured | Only MiniMax M3 measured |
| LMArena Text: software and IT | 1466 | Not measured | Only MiniMax M3 measured |
| LMArena WebDev: React | 1478 | Not measured | Only MiniMax M3 measured |
| LMArena WebDev: HTML | 1454 | Not measured | Only MiniMax M3 measured |
| LMArena WebDev: from an image | 1470 | Not measured | Only MiniMax M3 measured |
| Neither model has a result on the other 21 boards here | |||
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index10
| Benchmark | MiniMax M3 | Mistral Small 4 | Leader |
|---|---|---|---|
| AA Intelligence Index | 29.2 | 11.3 | MiniMax M3 |
| Epoch Capabilities Index | 146.9 | Not measured | Only MiniMax M3 measured |
Agentic workLike-for-likePublic boards, shown as published, kept out of the index20
| Benchmark | MiniMax M3 | Mistral Small 4 | Leader |
|---|---|---|---|
| AA Agentic Index | 29.5 | 0.8 | MiniMax M3 |
| Design Arena: agents building web apps | 1200 | Not measured | Only MiniMax M3 measured |
| Design Arena: agents building full-stack apps | 1197 | Not measured | Only MiniMax M3 measured |
| Design Arena: agents building mobile apps | 1169 | Not measured | Only MiniMax M3 measured |
| Design Arena: agents building Android apps | 1160 | Not measured | Only MiniMax M3 measured |
| Design Arena: agents building games | 1159 | Not measured | Only MiniMax M3 measured |
| Design Arena: HTML slides | 1086 | Not measured | Only MiniMax M3 measured |
| Design Arena: PowerPoint slides | 1169 | Not measured | Only MiniMax M3 measured |
| Design Arena: spreadsheets | 1042 | Not measured | Only MiniMax M3 measured |
| LiveBench research | 68.0% | Not measured | Only MiniMax M3 measured |
| LiveBench browser | 8.0% | Not measured | Only MiniMax M3 measured |
| LiveBench long agentic | 12.4% | Not measured | Only MiniMax M3 measured |
| Finance Agent (Vals AI) | 48.3% | Not measured | Only MiniMax M3 measured |
| Tax Agent Bench (Vals AI) | 20.6% | Not measured | Only MiniMax M3 measured |
| Legal Research Bench (Vals AI) | 29.8% | Not measured | Only MiniMax M3 measured |
| Excel Modeling Benchmark (Vals AI) | 47.8% | Not measured | Only MiniMax M3 measured |
| τ²-Bench Telecom | 88.9% | 41.2% | MiniMax M3 |
| APEX-Agents | 37.7% | Not measured | Only MiniMax M3 measured |
| OSWorld 2.0 | 4.6% | Not measured | Only MiniMax M3 measured |
| Vending-Bench 2 | $2,158 | Not measured | Only MiniMax M3 measured |
| Neither model has a result on the other 22 boards here | |||
ReasoningLike-for-likePublic boards, shown as published, kept out of the index20
| Benchmark | MiniMax M3 | Mistral Small 4 | Leader |
|---|---|---|---|
| GPQA Diamond | 92.7% | Not measured | Only MiniMax M3 measured |
| LiveBench reasoning | 74.5% | Not measured | Only MiniMax M3 measured |
| MMMU Pro (Vals AI) | 81.2% | Not measured | Only MiniMax M3 measured |
| GPQA Diamond (Artificial Analysis run) | 92.9% | 76.9% | MiniMax M3 |
| Humanity's Last Exam (Artificial Analysis run) | 39.0% | 9.9% | MiniMax M3 |
| GPQA Diamond (Epoch AI run) | 90.9% | Not measured | Only MiniMax M3 measured |
| SimpleBench | 45.8% | Not measured | Only MiniMax M3 measured |
| Chess puzzles | 13.0% | Not measured | Only MiniMax M3 measured |
| Mystery game puzzles | 8.0% | Not measured | Only MiniMax M3 measured |
| LMArena Text: hard prompts | 1447 | Not measured | Only MiniMax M3 measured |
| LMArena Text: expert | 1461 | Not measured | Only MiniMax M3 measured |
| LMArena Text: hard prompts in English | 1455 | Not measured | Only MiniMax M3 measured |
| LMArena Vision: diagrams | 1266 | Not measured | Only MiniMax M3 measured |
| LMArena Vision: homework | 1268 | Not measured | Only MiniMax M3 measured |
| Kagi LLM Benchmark | Not measured | 40.5% | Only Mistral Small 4 measured |
| Neither model has a result on the other 8 boards here | |||
MathDirectional onlyPublic boards, shown as published, kept out of the index4 boards, none shared
| Benchmark | MiniMax M3 | Mistral Small 4 | Leader |
|---|---|---|---|
| OTIS Mock AIME | 71.1% | Not measured | Only MiniMax M3 measured |
| ProofBench | 18.0% | Not measured | Only MiniMax M3 measured |
| LMArena Text: math | 1429 | Not measured | Only MiniMax M3 measured |
| LMArena Text: mathematical work | 1448 | Not measured | Only MiniMax M3 measured |
| Neither model has a result on the other 10 boards here | |||
KnowledgeDirectional onlyPublic boards, shown as published, kept out of the index10 boards, none shared
| Benchmark | MiniMax M3 | Mistral Small 4 | Leader |
|---|---|---|---|
| MMLU Pro | 84.2% | Not measured | Only MiniMax M3 measured |
| LegalBench (Vals AI) | 85.4% | Not measured | Only MiniMax M3 measured |
| CorpFin (Vals AI) | 68.1% | Not measured | Only MiniMax M3 measured |
| TaxEval (Vals AI) | 72.7% | Not measured | Only MiniMax M3 measured |
| MortgageTax (Vals AI) | 68.4% | Not measured | Only MiniMax M3 measured |
| MedCode (Vals AI) | 46.3% | Not measured | Only MiniMax M3 measured |
| LMArena Text: science | 1451 | Not measured | Only MiniMax M3 measured |
| LMArena Text: medicine and healthcare | 1439 | Not measured | Only MiniMax M3 measured |
| LMArena Text: legal and government | 1436 | Not measured | Only MiniMax M3 measured |
| LMArena Text: business and finance | 1434 | Not measured | Only MiniMax M3 measured |
| Neither model has a result on the other 10 boards here | |||
Web designDirectional onlyPublic boards, shown as published, kept out of the index9 boards, none shared
| Benchmark | MiniMax M3 | Mistral Small 4 | Leader |
|---|---|---|---|
| Design Arena: websites | 1257 | Not measured | Only MiniMax M3 measured |
| Design Arena: UI components | 1243 | Not measured | Only MiniMax M3 measured |
| Design Arena: game dev | 1211 | Not measured | Only MiniMax M3 measured |
| Design Arena: data viz | 1236 | Not measured | Only MiniMax M3 measured |
| Design Arena: 3D | 1209 | Not measured | Only MiniMax M3 measured |
| Design Arena: all code categories | 1246 | Not measured | Only MiniMax M3 measured |
| Design Arena: SVG | 1161 | Not measured | Only MiniMax M3 measured |
| Design Arena: ASCII art | 1161 | Not measured | Only MiniMax M3 measured |
| LiveBench design | 66.2% | Not measured | Only MiniMax M3 measured |
GeneralLike-for-likePublic boards, shown as published, kept out of the index40
| Benchmark | MiniMax M3 | Mistral Small 4 | Leader |
|---|---|---|---|
| LiveBench data analysis | 76.2% | Not measured | Only MiniMax M3 measured |
| Public Benefits Bench (Vals AI) | 64.1% | Not measured | Only MiniMax M3 measured |
| SAGE (Vals AI) | 50.6% | Not measured | Only MiniMax M3 measured |
| MedScribe (Vals AI) | 87.3% | Not measured | Only MiniMax M3 measured |
| Vals Multimodal Index | 60.0% | Not measured | Only MiniMax M3 measured |
| IFBench | 82.9% | 48.2% | MiniMax M3 |
| Long context reasoning (AA-LCR) | 83.0% | 49.7% | MiniMax M3 |
| DTBench | 78.9% | 70.9% | MiniMax M3 |
| LMCA | 33.7% | 20.6% | MiniMax M3 |
| Surface Evolver Bench | 55.0% | Not measured | Only MiniMax M3 measured |
| LMArena Text | 1432 | Not measured | Only MiniMax M3 measured |
| LMArena Text: instruction following | 1432 | Not measured | Only MiniMax M3 measured |
| LMArena Text: creative writing | 1403 | Not measured | Only MiniMax M3 measured |
| LMArena Vision | 1251 | Not measured | Only MiniMax M3 measured |
| LMArena Text: English | 1442 | Not measured | Only MiniMax M3 measured |
| LMArena Text: non-English | 1419 | Not measured | Only MiniMax M3 measured |
| LMArena Text: Chinese | 1465 | Not measured | Only MiniMax M3 measured |
| LMArena Text: German | 1430 | Not measured | Only MiniMax M3 measured |
| LMArena Text: Japanese | 1379 | Not measured | Only MiniMax M3 measured |
| LMArena Text: Korean | 1372 | Not measured | Only MiniMax M3 measured |
| LMArena Text: French | 1443 | Not measured | Only MiniMax M3 measured |
| LMArena Text: Spanish | 1432 | Not measured | Only MiniMax M3 measured |
| LMArena Text: Russian | 1427 | Not measured | Only MiniMax M3 measured |
| LMArena Text: Polish | 1425 | Not measured | Only MiniMax M3 measured |
| LMArena Text: multi-turn | 1443 | Not measured | Only MiniMax M3 measured |
| LMArena Text: longer queries | 1445 | Not measured | Only MiniMax M3 measured |
| LMArena Text: writing and language | 1418 | Not measured | Only MiniMax M3 measured |
| LMArena Text: entertainment and media | 1402 | Not measured | Only MiniMax M3 measured |
| LMArena Vision: English | 1263 | Not measured | Only MiniMax M3 measured |
| LMArena Vision: Chinese | 1311 | Not measured | Only MiniMax M3 measured |
| LMArena Vision: reading text | 1259 | Not measured | Only MiniMax M3 measured |
| LMArena Vision: humor | 1235 | Not measured | Only MiniMax M3 measured |
| LMArena Vision: recognizing things | 1197 | Not measured | Only MiniMax M3 measured |
| LMArena Vision: creative writing | 1234 | Not measured | Only MiniMax M3 measured |
| LMArena Document | 1435 | Not measured | Only MiniMax M3 measured |
| EQ-Bench 4 | 1150 | Not measured | Only MiniMax M3 measured |
| Neither model has a result on the other 12 boards here | |||
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | MiniMax M3 | Mistral Small 4 | Leader |
|---|---|---|---|
| Neither model has a result on 5 boards here | |||
What each workload costs
Three fixed token mixes at each model's listed price, so per-token rates become a bill.
Chat turn
1K input + 500 output tokens
Mistral Small 4 costs less
Repository review
50K input + 3K output tokens
Mistral Small 4 costs less
Agent loop
200K input + 10K output tokens
Mistral Small 4 costs less
Listed OpenRouter rates, without cache or batch discounts. A workload that does not fit a model's context window would need splitting.
Specification differences
The leading value in a row is marked; a row with a missing value names no leader. Our two rows are marked in red.
| Specification | MiniMax M3 | Mistral Small 4 |
|---|---|---|
| Maker | MiniMax | Mistral AI |
| Weights | Open weights | Open weights |
| Released | May 31, 2026 | March 16, 2026 |
| Context window | 1.05M | 262K |
| Input $/Mtok | $0.30 | $0.15 |
| Output $/Mtok | $1.20 | $0.60 |
| Coding index | 39.2 | Not benchmarked |
| Independent measurements behind it | 6 | 1 |
| Cost per solved task | $0.56 | Not computable |
| Source | OpenRouter listing | OpenRouter listing |
Questions
Is MiniMax M3 better than Mistral Small 4 for coding?
Only MiniMax M3 has a coding score, 39.2 on the Vibe Coding Index; Mistral Small 4 has none, so the two cannot be ranked on coding. Pick Mistral Small 4 for output tokens 2.0x cheaper ($0.60 vs $1.20 per million).
Which is cheaper, MiniMax M3 or Mistral Small 4?
MiniMax M3 costs $0.30 per million input tokens and $1.20 per million output tokens; Mistral Small 4 costs $0.15 and $0.60. Cost per solved task is known for MiniMax M3 only.
Which has the bigger context window, MiniMax M3 or Mistral Small 4?
MiniMax M3, with 1.05M tokens against 262K.
How many benchmarks is the MiniMax M3 vs Mistral Small 4 comparison based on?
14 benchmarks have a result for both models as of October 23, 2026. MiniMax M3's coding index rests on 6 independent benchmarks and Mistral Small 4's on 1. Where only one model has a result, the page says so instead of naming a winner.