Kimi K2 Thinking vs Mistral Medium 3.1
Kimi K2 Thinking leads Mistral Medium 3.1 on the Artificial Analysis Intelligence Index, 22.0 to 9.2, and Mistral Medium 3.1 costs 1.3x less per output token ($2.00 vs $2.50 per million). Updated October 21, 2026; 21 benchmarks have a result for both.
Kimi K2 Thinking or Mistral Medium 3.1: which to pick
Neither model has a coding score yet, so the two cannot be ranked on coding.
Pick Kimi K2 Thinking for a larger context window (262K against 131K tokens) and work measured by MGSM (Vals AI) or Design Arena: websites, where it leads.
- Shared results
- 21
- Kimi K2 Thinking only
- 9
- Mistral Medium 3.1 only
- 53
- Like-for-like categories
- 8 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
Neither has a coding score yet.
One side not measured
Intelligence
AA Intelligence Index
Kimi K2 Thinking
Kimi K2 Thinking leads, 22.0 against 9.2.
Same board, both measured
Agentic work
AA Agentic Index
No call
Only Mistral Medium 3.1 has a AA Agentic Index result, 1.9.
Directional only
Reasoning
GPQA Diamond
No call
Only Kimi K2 Thinking has a GPQA Diamond result, 78.5%.
Directional only
Math
AIME
Kimi K2 Thinking
Kimi K2 Thinking leads, 85.4% against 42.3%.
Same board, both measured
Knowledge
MMLU Pro
Kimi K2 Thinking
Kimi K2 Thinking leads, 81.1% against 75.3%.
Same board, both measured
Web design
Design Arena: websites
Mistral Medium 3.1
Mistral Medium 3.1 leads, 1133 against 1113.
Same board, both measured
Long documents
Listed context window
Kimi K2 Thinking
Kimi K2 Thinking takes 262K tokens against 131K.
Listed by OpenRouter
Everyday cost
1K input + 500 output tokens
Mistral Medium 3.1
Kimi K2 Thinking $0.0019, Mistral Medium 3.1 $0.0014 per turn.
List prices
Cost per solved task
One SWE-bench Verified task, solved
No call
Only Kimi K2 Thinking has the number, $0.61.
Directional only
At a glance
- Coding index
- Kimi K2 Thinking not benchmarked, Mistral Medium 3.1 not benchmarked.
- Widest category
- Agentic work, 52.3 points apart on average over 1 shared benchmark.
- Widest single gap
- AA Math Index: Kimi K2 Thinking 94.7 against Mistral Medium 3.1 38.3, 56.4 points apart on a 0 to 100 scale.
- Output price
- Kimi K2 Thinking $2.50 against Mistral Medium 3.1 $2.00 per million output tokens, 1.3x apart.
- Context window
- Kimi K2 Thinking takes 262K tokens, Mistral Medium 3.1 131K.
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 | Kimi K2 Thinking | Bars | Mistral Medium 3.1 | Ties |
|---|---|---|---|---|
| Coding | 4 | 0 | ||
| Intelligence | 1 | 0 | ||
| Agentic work | 1 | 0 | ||
| Reasoning | 2 | 0 | ||
| Math | 3 | 1 | ||
| Knowledge | 6 | 0 | ||
| Web design | 0 | 1 | ||
| General | 2 | 0 | ||
| All shared boards | 19 | 2 |
CodingLike-for-likeSix of these boards feed the Vibe Coding Index40
| Benchmark | Kimi K2 Thinking | Mistral Medium 3.1 | Leader |
|---|---|---|---|
| SWE-bench Verified | 63.4% | Not measured | Only Kimi K2 Thinking measured |
| AA Coding Index | Not measured | 20.5 | Only Mistral Medium 3.1 measured |
| LiveCodeBench (Vals AI) | 63.1% | 44.8% | Kimi K2 Thinking |
| SWE-bench Verified (Vals AI) | 60.2% | Not measured | Only Kimi K2 Thinking measured |
| LiveCodeBench (Artificial Analysis run) | 85.3% | 40.6% | Kimi K2 Thinking |
| SciCode | Not measured | 32.4% | Only Mistral Medium 3.1 measured |
| Terminal-Bench Hard (Artificial Analysis run) | 31.1% | 10.6% | Kimi K2 Thinking |
| Terminal-Bench | 35.7% | Not measured | Only Kimi K2 Thinking measured |
| WeirdML | 42.8% | Not measured | Only Kimi K2 Thinking measured |
| ALE-Bench | 597.5 | 210.2 | Kimi K2 Thinking |
| LMArena Text: coding | Not measured | 1433 | Only Mistral Medium 3.1 measured |
| LMArena Text: software and IT | Not measured | 1436 | Only Mistral Medium 3.1 measured |
| Neither model has a result on the other 26 boards here | |||
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index10
| Benchmark | Kimi K2 Thinking | Mistral Medium 3.1 | Leader |
|---|---|---|---|
| AA Intelligence Index | 22.0 | 9.2 | Kimi K2 Thinking |
| Epoch Capabilities Index | 145.8 | Not measured | Only Kimi K2 Thinking measured |
Agentic workLike-for-likePublic boards, shown as published, kept out of the index10
| Benchmark | Kimi K2 Thinking | Mistral Medium 3.1 | Leader |
|---|---|---|---|
| AA Agentic Index | Not measured | 1.9 | Only Mistral Medium 3.1 measured |
| τ²-Bench Telecom | 93.0% | 40.6% | Kimi K2 Thinking |
| METR time horizons | 59.2% | Not measured | Only Kimi K2 Thinking measured |
| Neither model has a result on the other 36 boards here | |||
ReasoningLike-for-likePublic boards, shown as published, kept out of the index20
| Benchmark | Kimi K2 Thinking | Mistral Medium 3.1 | Leader |
|---|---|---|---|
| GPQA Diamond | 78.5% | Not measured | Only Kimi K2 Thinking measured |
| MMMU Pro (Vals AI) | Not measured | 63.0% | Only Mistral Medium 3.1 measured |
| GPQA Diamond (Artificial Analysis run) | 83.8% | 58.8% | Kimi K2 Thinking |
| Humanity's Last Exam (Artificial Analysis run) | 23.8% | 4.7% | Kimi K2 Thinking |
| LMArena Text: hard prompts | Not measured | 1426 | Only Mistral Medium 3.1 measured |
| LMArena Text: expert | Not measured | 1408 | Only Mistral Medium 3.1 measured |
| LMArena Text: hard prompts in English | Not measured | 1435 | Only Mistral Medium 3.1 measured |
| LMArena Vision: diagrams | Not measured | 1175 | Only Mistral Medium 3.1 measured |
| LMArena Vision: homework | Not measured | 1165 | Only Mistral Medium 3.1 measured |
| Neither model has a result on the other 12 boards here | |||
MathLike-for-likePublic boards, shown as published, kept out of the index31
| Benchmark | Kimi K2 Thinking | Mistral Medium 3.1 | Leader |
|---|---|---|---|
| AIME | 85.4% | 42.3% | Kimi K2 Thinking |
| MATH 500 (Vals AI) | Not measured | 87.0% | Only Mistral Medium 3.1 measured |
| MGSM (Vals AI) | 90.2% | 90.9% | Mistral Medium 3.1 |
| AA Math Index | 94.7 | 38.3 | Kimi K2 Thinking |
| AIME 2025 (Artificial Analysis run) | 94.7% | 38.3% | Kimi K2 Thinking |
| LMArena Text: math | Not measured | 1408 | Only Mistral Medium 3.1 measured |
| LMArena Text: mathematical work | Not measured | 1412 | Only Mistral Medium 3.1 measured |
| Neither model has a result on the other 6 boards here | |||
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index60
| Benchmark | Kimi K2 Thinking | Mistral Medium 3.1 | Leader |
|---|---|---|---|
| MMLU Pro | 81.1% | 75.3% | Kimi K2 Thinking |
| LegalBench (Vals AI) | 80.2% | 62.0% | Kimi K2 Thinking |
| CaseLaw (Vals AI) | 65.7% | Not measured | Only Kimi K2 Thinking measured |
| CorpFin (Vals AI) | 60.6% | 50.7% | Kimi K2 Thinking |
| TaxEval (Vals AI) | 71.7% | 70.3% | Kimi K2 Thinking |
| MortgageTax (Vals AI) | Not measured | 36.5% | Only Mistral Medium 3.1 measured |
| MedQA (Vals AI) | 92.6% | 78.2% | Kimi K2 Thinking |
| MMLU Pro (Artificial Analysis run) | 84.8% | 68.3% | Kimi K2 Thinking |
| LMArena Text: science | Not measured | 1444 | Only Mistral Medium 3.1 measured |
| LMArena Text: medicine and healthcare | Not measured | 1449 | Only Mistral Medium 3.1 measured |
| LMArena Text: legal and government | Not measured | 1439 | Only Mistral Medium 3.1 measured |
| LMArena Text: business and finance | Not measured | 1418 | Only Mistral Medium 3.1 measured |
| Neither model has a result on the other 5 boards here | |||
Web designLike-for-likePublic boards, shown as published, kept out of the index01
| Benchmark | Kimi K2 Thinking | Mistral Medium 3.1 | Leader |
|---|---|---|---|
| Design Arena: websites | 1113 | 1133 | Mistral Medium 3.1 |
| Design Arena: UI components | Not measured | 1100 | Only Mistral Medium 3.1 measured |
| Design Arena: game dev | Not measured | 1074 | Only Mistral Medium 3.1 measured |
| Design Arena: data viz | Not measured | 1144 | Only Mistral Medium 3.1 measured |
| Design Arena: 3D | Not measured | 1081 | Only Mistral Medium 3.1 measured |
| Design Arena: all code categories | Not measured | 1120 | Only Mistral Medium 3.1 measured |
| Design Arena: SVG | Not measured | 993 | Only Mistral Medium 3.1 measured |
| Design Arena: ASCII art | Not measured | 997 | Only Mistral Medium 3.1 measured |
| Neither model has a result on the other 1 board here | |||
GeneralLike-for-likePublic boards, shown as published, kept out of the index20
| Benchmark | Kimi K2 Thinking | Mistral Medium 3.1 | Leader |
|---|---|---|---|
| IFBench | 68.1% | 39.8% | Kimi K2 Thinking |
| Long context reasoning (AA-LCR) | 72.0% | 42.7% | Kimi K2 Thinking |
| DTBench | Not measured | 67.5% | Only Mistral Medium 3.1 measured |
| LMCA | Not measured | 19.1% | Only Mistral Medium 3.1 measured |
| CL-bench | 17.6% | Not measured | Only Kimi K2 Thinking measured |
| LMArena Text | Not measured | 1424 | Only Mistral Medium 3.1 measured |
| LMArena Text: instruction following | Not measured | 1398 | Only Mistral Medium 3.1 measured |
| LMArena Text: creative writing | Not measured | 1391 | Only Mistral Medium 3.1 measured |
| LMArena Vision | Not measured | 1172 | Only Mistral Medium 3.1 measured |
| LMArena Text: English | Not measured | 1438 | Only Mistral Medium 3.1 measured |
| LMArena Text: non-English | Not measured | 1408 | Only Mistral Medium 3.1 measured |
| LMArena Text: Chinese | Not measured | 1447 | Only Mistral Medium 3.1 measured |
| LMArena Text: German | Not measured | 1432 | Only Mistral Medium 3.1 measured |
| LMArena Text: Japanese | Not measured | 1379 | Only Mistral Medium 3.1 measured |
| LMArena Text: Korean | Not measured | 1380 | Only Mistral Medium 3.1 measured |
| LMArena Text: French | Not measured | 1460 | Only Mistral Medium 3.1 measured |
| LMArena Text: Spanish | Not measured | 1433 | Only Mistral Medium 3.1 measured |
| LMArena Text: Russian | Not measured | 1411 | Only Mistral Medium 3.1 measured |
| LMArena Text: Polish | Not measured | 1430 | Only Mistral Medium 3.1 measured |
| LMArena Text: multi-turn | Not measured | 1418 | Only Mistral Medium 3.1 measured |
| LMArena Text: longer queries | Not measured | 1406 | Only Mistral Medium 3.1 measured |
| LMArena Text: writing and language | Not measured | 1395 | Only Mistral Medium 3.1 measured |
| LMArena Text: entertainment and media | Not measured | 1390 | Only Mistral Medium 3.1 measured |
| LMArena Vision: English | Not measured | 1179 | Only Mistral Medium 3.1 measured |
| LMArena Vision: Chinese | Not measured | 1190 | Only Mistral Medium 3.1 measured |
| LMArena Vision: reading text | Not measured | 1175 | Only Mistral Medium 3.1 measured |
| LMArena Vision: humor | Not measured | 1168 | Only Mistral Medium 3.1 measured |
| LMArena Vision: captioning | Not measured | 1145 | Only Mistral Medium 3.1 measured |
| LMArena Vision: recognizing things | Not measured | 1147 | Only Mistral Medium 3.1 measured |
| LMArena Vision: creative writing | Not measured | 1200 | Only Mistral Medium 3.1 measured |
| Neither model has a result on the other 16 boards here | |||
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | Kimi K2 Thinking | Mistral Medium 3.1 | 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 Medium 3.1 costs less
Repository review
50K input + 3K output tokens
Mistral Medium 3.1 costs less
Agent loop
200K input + 10K output tokens
Mistral Medium 3.1 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 | MoonshotAI Kimi K2 Thinking | Mistral Medium 3.1 |
|---|---|---|
| Maker | Moonshot AI | Mistral AI |
| Weights | Open weights | No public weights listed |
| Released | November 6, 2025 | August 13, 2025 |
| Context window | 262K | 131K |
| Input $/Mtok | $0.60 | $0.40 |
| Output $/Mtok | $2.50 | $2.00 |
| Coding index | Not benchmarked | Not benchmarked |
| Independent measurements behind it | 1 | 0 |
| Cost per solved task | $0.61 | Not computable |
| Source | OpenRouter listing | OpenRouter listing |
Questions
Is Kimi K2 Thinking better than Mistral Medium 3.1 for coding?
Neither model has a coding score yet, so the two cannot be ranked on coding. Pick Kimi K2 Thinking for a larger context window (262K against 131K tokens) and work measured by MGSM (Vals AI) or Design Arena: websites, where it leads.
Which is cheaper, Kimi K2 Thinking or Mistral Medium 3.1?
Kimi K2 Thinking costs $0.60 per million input tokens and $2.50 per million output tokens; Mistral Medium 3.1 costs $0.40 and $2.00. Cost per solved task is known for Kimi K2 Thinking only.
Which has the bigger context window, Kimi K2 Thinking or Mistral Medium 3.1?
Kimi K2 Thinking, with 262K tokens against 131K.
How many benchmarks is the Kimi K2 Thinking vs Mistral Medium 3.1 comparison based on?
21 benchmarks have a result for both models as of October 21, 2026. Kimi K2 Thinking's coding index rests on 1 independent benchmark and Mistral Medium 3.1's on 0. Where only one model has a result, the page says so instead of naming a winner.