Model comparison
MiniMax-M2.5 vs Mistral Small
MiniMax-M2.5 is the stronger model overall, scoring 38.3 to 33.4 on the Noometry Index. Mistral Small costs 2.0× less per token, which makes it the better buy when MiniMax-M2.5's lead doesn't matter for your workload.
Last verified . 20 shared benchmarks.
MiniMax
Mistral AISummary
- They share 20 benchmarks with published results for both. MiniMax-M2.5 scores higher in 7 categories and Mistral Small in 2 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in coding, where MiniMax-M2.5 leads 48.1 to 34.0.
- The biggest single-benchmark swing is Kagi LLM Benchmark: 55.2% for MiniMax-M2.5 and 37.8% for Mistral Small.
- Mistral Small is cheaper at $0.15 / $0.60 per million input/output tokens, against $0.30 / $1.20 for MiniMax-M2.5.
- Mistral Small accepts more context: 262K tokens versus 205K.
Side by side
| MiniMax-M2.5 | Mistral Small | |
|---|---|---|
| Provider | MiniMax | Mistral AI |
| Noometry Index | 38.3 | 33.4 |
| Released | 2026-02-12 | 2024-02-26 |
| Weights | Open | Open |
| Context window | 205K | 262K |
| Max output | 131K | 256K |
| Input $ / M tokens | $0.30 | $0.15 |
| Output $ / M tokens | $1.20 | $0.60 |
| Results tracked | 33 | 39 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding MiniMax-M2.5 leads
MiniMax-M2.5: 48.1 (#58), Mistral Small: 34.0 (#247)
| Benchmark | MiniMax-M2.5 | Mistral Small |
|---|---|---|
| LMArena Coding | 1381 | 1362 |
| ALE-Bench | 618.17 | 497.62 |
| SWE-bench Verified (bash only) | 75.8% | — |
| LMArena WebDev | 1387 | — |
| SWE-bench Multilingual | 68.3% | — |
| SciCode | — | 26.5% |
| BigCodeBench Instruct | — | 36.1% |
| LiveBench Coding | — | 36.2% |
| BigCodeBench Complete | — | 46.6% |
Agentic & Tool Use MiniMax-M2.5 leads
MiniMax-M2.5: 30.4 (#77), Mistral Small: 28.1 (#93)
| Benchmark | MiniMax-M2.5 | Mistral Small |
|---|---|---|
| Terminal-Bench | 42.7% | — |
| Berkeley Function Calling Leaderboard | — | 37.1% |
| Vending-Bench 2 | -23.16 | — |
Reasoning Mistral Small leads
MiniMax-M2.5: 17.5 (#292), Mistral Small: 19.8 (#250)
| Benchmark | MiniMax-M2.5 | Mistral Small |
|---|---|---|
| Kagi LLM Benchmark | 55.2% | 37.8% |
| LMArena Hard Prompts | 1372 | 1335 |
| ARC-AGI-2 | 4.9% | — |
| NYT Connections (extended) | 16.8% | — |
| ARC-AGI-1 | 63.7% | — |
| CritPt | — | 0% |
| LiveBench Reasoning | — | 44.8% |
| DTBench | — | 70.9% |
| LiveBench Data Analysis | — | 53.7% |
| LMCA | — | 20.6% |
| Epoch Capabilities Index | 146.68 | — |
| LiveBench | — | 44% |
Math MiniMax-M2.5 leads
MiniMax-M2.5: 26.9 (#253), Mistral Small: 16.4 (#293)
| Benchmark | MiniMax-M2.5 | Mistral Small |
|---|---|---|
| LMArena Math | 1378 | 1341 |
| OTIS Mock AIME 2024-2025 | — | 5.8% |
| ProofBench | 4% | — |
| LiveBench Math | — | 39.9% |
| MATH Level 5 | — | 46.8% |
Knowledge MiniMax-M2.5 leads
MiniMax-M2.5: 39.2 (#135), Mistral Small: 31.0 (#222)
| Benchmark | MiniMax-M2.5 | Mistral Small |
|---|---|---|
| Vectara Hallucination Rate | 9.1% | 5.1% |
| LMArena Expert | 1379 | 1291 |
| GPQA Diamond | — | 47.5% |
| MMLU | — | 68.7% |
Multimodal Not comparable
MiniMax-M2.5: —, Mistral Small: 33.5 (#96)
| Benchmark | MiniMax-M2.5 | Mistral Small |
|---|---|---|
| LMArena Vision | — | 1142 |
Multilingual MiniMax-M2.5 leads
MiniMax-M2.5: 47.1 (#152), Mistral Small: 45.5 (#169)
| Benchmark | MiniMax-M2.5 | Mistral Small |
|---|---|---|
| LMArena Non-English | 1338 | 1315 |
| LMArena Chinese | 1393 | 1340 |
| LMArena French | 1362 | 1337 |
| LMArena German | 1362 | 1340 |
| LMArena Japanese | 1171 | 1275 |
| LMArena Korean | 1232 | 1259 |
| LMArena Russian | 1358 | 1324 |
| LMArena Spanish | 1354 | 1346 |
Instruction Following MiniMax-M2.5 leads
MiniMax-M2.5: 71.5 (#148), Mistral Small: 66.4 (#209)
| Benchmark | MiniMax-M2.5 | Mistral Small |
|---|---|---|
| LMArena Instruction Following | 1353 | 1310 |
| LiveBench Instruction Following | — | 63.7% |
Long Context Mistral Small leads
MiniMax-M2.5: 37.5 (#216), Mistral Small: 40.4 (#156)
| Benchmark | MiniMax-M2.5 | Mistral Small |
|---|---|---|
| LMArena Longer Query | 1366 | 1327 |
| CL-bench | 11.4% | — |
| CL-bench Life | 6.3% | — |
Writing & Preference MiniMax-M2.5 leads
MiniMax-M2.5: 53.9 (#153), Mistral Small: 52.5 (#171)
| Benchmark | MiniMax-M2.5 | Mistral Small |
|---|---|---|
| LMArena Text | 1359 | 1338 |
| LMArena Creative Writing | 1331 | 1305 |
| LMArena Multi-Turn | 1364 | 1344 |
| EQ-Bench Creative Writing | 1361 | — |
| LiveBench Language | — | 30.5% |
Frequently asked questions
Is MiniMax-M2.5 better than Mistral Small?
MiniMax-M2.5 is the stronger model overall, scoring 38.3 to 33.4 on the Noometry Index. Mistral Small costs 2.0× less per token, which makes it the better buy when MiniMax-M2.5's lead doesn't matter for your workload.
Which is cheaper, MiniMax-M2.5 or Mistral Small?
Mistral Small is cheaper. It lists at $0.15 per million input tokens and $0.60 per million output tokens; MiniMax-M2.5 lists at $0.30 and $1.20.
Is MiniMax-M2.5 or Mistral Small better for coding?
MiniMax-M2.5 scores higher on coding benchmarks: 48.1 versus 34.0 in the Noometry coding category.
Which has the bigger context window?
Mistral Small does, with 262K tokens against 205K.
How many benchmarks do MiniMax-M2.5 and Mistral Small share?
20 benchmarks have published results for both models. MiniMax-M2.5 has 33 scored results on Noometry and Mistral Small has 39.