Model comparison
Devstral Small 2505 vs Kimi K2 Thinking Turbo
Kimi K2 Thinking Turbo is the stronger model overall, scoring 45.4 to 33.9 on the Noometry Index.
Last verified . 0 shared benchmarks.
Mistral AISummary
- The widest gap is in reasoning, where Kimi K2 Thinking Turbo leads 31.4 to 19.1.
Side by side
| Devstral Small 2505 | Kimi K2 Thinking Turbo | |
|---|---|---|
| Provider | Mistral AI | Moonshot AI |
| Noometry Index | 33.9 | 45.4 |
| Released | 2025-05-07 | 2025-11-06 |
| Weights | Open | Open |
| Context window | 128K | — |
| Max output | 128K | — |
| Input $ / M tokens | $0.10 | — |
| Output $ / M tokens | $0.30 | — |
| Results tracked | 4 | 21 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Too close to call
Devstral Small 2505: 38.6 (#160), Kimi K2 Thinking Turbo: 38.1 (#166)
| Benchmark | Devstral Small 2505 | Kimi K2 Thinking Turbo |
|---|---|---|
| SWE-bench Verified (bash only) | 56.4% | — |
| LMArena WebDev | — | 1322 |
| SciCode | 28.8% | — |
| LMArena Coding | — | 1454 |
Reasoning Kimi K2 Thinking Turbo leads
Devstral Small 2505: 19.1 (#243), Kimi K2 Thinking Turbo: 31.4 (#72)
| Benchmark | Devstral Small 2505 | Kimi K2 Thinking Turbo |
|---|---|---|
| Kagi LLM Benchmark | 37.7% | — |
| CritPt | 0% | — |
| Chess Puzzles | — | 20% |
| LMArena Hard Prompts | — | 1428 |
Math Not comparable
Devstral Small 2505: —, Kimi K2 Thinking Turbo: 45.5 (#62)
| Benchmark | Devstral Small 2505 | Kimi K2 Thinking Turbo |
|---|---|---|
| OTIS Mock AIME 2024-2025 | — | 83.1% |
| LMArena Math | — | 1429 |
Knowledge Not comparable
Devstral Small 2505: —, Kimi K2 Thinking Turbo: 50.9 (#64)
| Benchmark | Devstral Small 2505 | Kimi K2 Thinking Turbo |
|---|---|---|
| GPQA Diamond | — | 84.2% |
| LMArena Expert | — | 1439 |
Multilingual Not comparable
Devstral Small 2505: —, Kimi K2 Thinking Turbo: 51.6 (#104)
| Benchmark | Devstral Small 2505 | Kimi K2 Thinking Turbo |
|---|---|---|
| LMArena Non-English | — | 1398 |
| LMArena Chinese | — | 1457 |
| LMArena French | — | 1425 |
| LMArena German | — | 1390 |
| LMArena Japanese | — | 1356 |
| LMArena Korean | — | 1329 |
| LMArena Russian | — | 1390 |
| LMArena Spanish | — | 1407 |
Instruction Following Not comparable
Devstral Small 2505: —, Kimi K2 Thinking Turbo: 74.2 (#103)
| Benchmark | Devstral Small 2505 | Kimi K2 Thinking Turbo |
|---|---|---|
| LMArena Instruction Following | — | 1403 |
Long Context Not comparable
Devstral Small 2505: —, Kimi K2 Thinking Turbo: 43.2 (#99)
| Benchmark | Devstral Small 2505 | Kimi K2 Thinking Turbo |
|---|---|---|
| LMArena Longer Query | — | 1415 |
Writing & Preference Not comparable
Devstral Small 2505: —, Kimi K2 Thinking Turbo: 60.2 (#99)
| Benchmark | Devstral Small 2505 | Kimi K2 Thinking Turbo |
|---|---|---|
| LMArena Text | — | 1415 |
| LMArena Creative Writing | — | 1373 |
| LMArena Multi-Turn | — | 1414 |
Frequently asked questions
Is Devstral Small 2505 better than Kimi K2 Thinking Turbo?
Kimi K2 Thinking Turbo is the stronger model overall, scoring 45.4 to 33.9 on the Noometry Index.
Is Devstral Small 2505 or Kimi K2 Thinking Turbo better for coding?
They score almost the same on coding (38.6 vs 38.1); test both on your own repository before choosing.
How many benchmarks do Devstral Small 2505 and Kimi K2 Thinking Turbo share?
0 benchmarks have published results for both models. Devstral Small 2505 has 4 scored results on Noometry and Kimi K2 Thinking Turbo has 21.