Compare / Gemma 4 31B vs Mistral Medium 3
Gemma 4 31B vs Mistral Medium 3
Which is cheaper, which is more capable, and which does better at each kind of work, from 16 results on 3 sources that measured both models.
Gemma 4 31B or Mistral Medium 3? The short answer
- Gemma 4 31B is cheaper: $0.21 against $0.80 per million tokens, 3.9 times less.
- Gemma 4 31B scores higher on the Artificial Analysis Intelligence Index: 14.7 against 9.0.
- On our business benchmarks, Gemma 4 31B ranks higher for reasoning and knowledge and writing.
- On our business benchmarks, Mistral Medium 3 ranks higher for agents and tool use.
At a glance
| Gemma 4 31B | Mistral Medium 3 | |
|---|---|---|
| Price per million tokens | $0.14 in · $0.40 out | $0.40 in · $2.00 out |
| Blended (3 in : 1 out) | $0.21 | $0.80 |
| Cheapest host, blended | $0.15 (DeepInfra, FP4) | $0.80 (Mistral) |
| Intelligence Index | 14.7 | 9.0 |
| Spring Prompt overall | – | – |
| Output speed, tokens a second | 44 | – |
| Context window | 262,144 tokens | 131,072 tokens |
| Weights | Open | Closed |
| Developer based in | the US | France |
Prices are the developer's list price where we have read it, otherwise the typical host on OpenRouter; see every model's price. Intelligence and speed from Artificial Analysis.
Which is better for what
Each model's rank among models available today, for each kind of work, from our own benchmarks and licensed sources. The better rank is in green.
| Use case | Gemma 4 31B | Mistral Medium 3 |
|---|---|---|
| Agents and tool useMulti-step tasks with tools: support desks, coding agents, function calls | 96 of 161 | 81 of 161 |
| Reasoning and knowledgeHard questions across science, maths and general knowledge | 89 of 195 | 160 of 195 |
| WritingWhat people prefer in blind comparisons, and judged writing quality | 72 of 187 | 127 of 187 |
| Sticking to the factsSummarising without inventing things, and factual answers | 89 of 153 | not measured |
Every shared result
| Benchmark | Metric | Gemma 4 31B | Mistral Medium 3 |
|---|---|---|---|
| Arena (formerly LMArena)reported | Business, management and financeArena rating | 1,449 | 1,387 |
| Arena (formerly LMArena)reported | Creative writingArena rating | 1,419 | 1,364 |
| Arena (formerly LMArena)reported | Expert promptsArena rating | 1,499 | 1,379 |
| Arena (formerly LMArena)reported | Instruction followingArena rating | 1,453 | 1,366 |
| Arena (formerly LMArena)reported | OverallArena rating | 1,453 | 1,387 |
| Arena (formerly LMArena)reported | Writing, literature and languageArena rating | 1,435 | 1,370 |
| Artificial Analysisreported | Artificial Analysis Intelligence Indexindex score | 14.7 | 9.00 |
| Artificial Analysisreported | GPQA Diamond% of questions | 85.7% | 57.8% |
| Artificial Analysisreported | Humanity's Last Exam% of questions | 23.6% | 4.1% |
| Artificial Analysisreported | IFBench% of instructions | 75.6% | 39.3% |
| Artificial Analysisreported | Long-context reasoning (AA-LCR)% of questions | 69.7% | 31.3% |
| Artificial Analysisreported | Terminal-Bench Hard% of tasks | 36.4% | 3.8% |
| Artificial Analysisreported | Τ²-bench telecom% of tasks | 65.5% | 24.3% |
| UGI Leaderboardreported | Requested-length error% off the requested word count | 5.0% | 20.0% |
| UGI Leaderboardreported | Style adherencescore from 0 to 1 | 0.36 | 0.35 |
| UGI Leaderboardreported | Writing scorescore out of 100 | 43.1 | 32.6 |
Bold green marks the better value on that metric. Where a source reports ranges that overlap, the difference may not be meaningful; see the benchmark page for ranges.
Quick answers
Is Gemma 4 31B better than Mistral Medium 3?
It depends on the task. Among models available today, on our benchmarks and licensed sources, Gemma 4 31B ranks higher for reasoning and knowledge and writing; Mistral Medium 3 ranks higher for agents and tool use.
Which is cheaper, Gemma 4 31B or Mistral Medium 3?
Gemma 4 31B costs $0.21 per million tokens (three input to one output), against $0.80 for Mistral Medium 3.
Run Gemma 4 31B and Mistral Medium 3 on your own prompt
Benchmarks aren't your data. Try both side by side in the playground, with the cost of every answer.