Model comparison
GPT-5.4 mini vs Mistral Medium
GPT-5.4 mini is the stronger model overall, scoring 45.0 to 36.3 on the Noometry Index.
Last verified . 31 shared benchmarks.
Mistral AISummary
- They share 31 benchmarks with published results for both. GPT-5.4 mini scores higher in 9 categories and Mistral Medium in 1 category; 7 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where GPT-5.4 mini leads 51.5 to 25.0.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 88.9% for GPT-5.4 mini and 32.2% for Mistral Medium.
- GPT-5.4 mini is cheaper at $0.75 / $4.50 per million input/output tokens, against $1.50 / $7.50 for Mistral Medium.
- GPT-5.4 mini accepts more context: 400K tokens versus 262K.
- Mistral Medium has downloadable open weights; the other is API-only.
Side by side
| GPT-5.4 mini | Mistral Medium | |
|---|---|---|
| Provider | OpenAI | Mistral AI |
| Noometry Index | 45.0 | 36.3 |
| Released | 2026-03-17 | 2023-12-11 |
| Weights | Proprietary | Open |
| Context window | 400K | 262K |
| Max output | 128K | 262K |
| Input $ / M tokens | $0.75 | $1.50 |
| Output $ / M tokens | $4.50 | $7.50 |
| Results tracked | 46 | 36 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-5.4 mini leads
GPT-5.4 mini: 45.2 (#72), Mistral Medium: 34.2 (#243)
| Benchmark | GPT-5.4 mini | Mistral Medium |
|---|---|---|
| FrontierCode | 27% | 8% |
| SciCode | 49.9% | 40.2% |
| WeirdML | 60.3% | 43.7% |
| LMArena Coding | 1438 | 1434 |
| ALE-Bench | 1,189 | 763.98 |
| LMArena WebDev | 1397 | — |
Agentic & Tool Use GPT-5.4 mini leads
GPT-5.4 mini: 29.9 (#81), Mistral Medium: 28.3 (#90)
| Benchmark | GPT-5.4 mini | Mistral Medium |
|---|---|---|
| Berkeley Function Calling Leaderboard | — | 37.7% |
| DeepResearch Bench | 36.3% | — |
Reasoning GPT-5.4 mini leads
GPT-5.4 mini: 30.4 (#85), Mistral Medium: 24.0 (#167)
| Benchmark | GPT-5.4 mini | Mistral Medium |
|---|---|---|
| Kagi LLM Benchmark | 37.9% | 50% |
| CritPt | 10% | 0% |
| LMArena Hard Prompts | 1424 | 1426 |
| DTBench | 80% | 75.5% |
| LMCA | 40.8% | 26.1% |
| ARC-AGI-2 | 18.9% | — |
| NYT Connections (extended) | 61.8% | — |
| ARC-AGI-1 | 63.7% | — |
| Chess Puzzles | 24% | — |
| Thematic Generalization | 61.7% | — |
| Mystery Game Puzzles | 11% | — |
| Surface Evolver Bench | — | 26.9% |
| Epoch Capabilities Index | 148.84 | — |
| ForecastBench | 57 | — |
Math GPT-5.4 mini leads
GPT-5.4 mini: 45.5 (#75), Mistral Medium: 28.1 (#245)
| Benchmark | GPT-5.4 mini | Mistral Medium |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 88.9% | 32.2% |
| ProofBench | 21% | 9% |
| LMArena Math | 1419 | 1408 |
| FrontierMath (Feb 2025 set) | 28.3% | 0.3% |
| FrontierMath (Tiers 1-3) | 51.2% | — |
| FrontierMath Tier 4 | 9.8% | — |
| MATH Level 5 | — | 81.6% |
| FrontierMath Tier 4 (v1) | 2.1% | — |
Knowledge GPT-5.4 mini leads
GPT-5.4 mini: 51.5 (#67), Mistral Medium: 25.0 (#265)
| Benchmark | GPT-5.4 mini | Mistral Medium |
|---|---|---|
| GPQA Diamond | 86.9% | 59.5% |
| Vectara Hallucination Rate | 5.5% | 22.7% |
| LMArena Expert | 1435 | 1408 |
| Humanity's Last Exam | — | 4.5% |
| SimpleQA Verified | 29.4% | — |
Multimodal GPT-5.4 mini leads
GPT-5.4 mini: 39.7 (#56), Mistral Medium: 35.3 (#88)
| Benchmark | GPT-5.4 mini | Mistral Medium |
|---|---|---|
| LMArena Vision | 1245 | 1172 |
Multilingual Too close to call
GPT-5.4 mini: 51.9 (#96), Mistral Medium: 52.1 (#91)
| Benchmark | GPT-5.4 mini | Mistral Medium |
|---|---|---|
| LMArena Non-English | 1405 | 1408 |
| LMArena Chinese | 1446 | 1447 |
| LMArena French | 1440 | 1459 |
| LMArena German | 1409 | 1432 |
| LMArena Japanese | 1374 | 1378 |
| LMArena Korean | 1368 | 1380 |
| LMArena Russian | 1417 | 1411 |
| LMArena Spanish | 1405 | 1433 |
Instruction Following Too close to call
GPT-5.4 mini: 74.1 (#102), Mistral Medium: 73.7 (#116)
| Benchmark | GPT-5.4 mini | Mistral Medium |
|---|---|---|
| LMArena Instruction Following | 1405 | 1398 |
Long Context Too close to call
GPT-5.4 mini: 43.0 (#112), Mistral Medium: 42.9 (#114)
| Benchmark | GPT-5.4 mini | Mistral Medium |
|---|---|---|
| LMArena Longer Query | 1407 | 1406 |
Writing & Preference GPT-5.4 mini leads
GPT-5.4 mini: 64.0 (#58), Mistral Medium: 60.0 (#103)
| Benchmark | GPT-5.4 mini | Mistral Medium |
|---|---|---|
| LMArena Text | 1412 | 1424 |
| LMArena Creative Writing | 1370 | 1391 |
| LMArena Multi-Turn | 1429 | 1418 |
| Short-Story Creative Writing | — | 77.3% |
| EQ-Bench Creative Writing | 1665 | — |
Frequently asked questions
Is GPT-5.4 mini better than Mistral Medium?
GPT-5.4 mini is the stronger model overall, scoring 45.0 to 36.3 on the Noometry Index.
Which is cheaper, GPT-5.4 mini or Mistral Medium?
GPT-5.4 mini is cheaper. It lists at $0.75 per million input tokens and $4.50 per million output tokens; Mistral Medium lists at $1.50 and $7.50.
Is GPT-5.4 mini or Mistral Medium better for coding?
GPT-5.4 mini scores higher on coding benchmarks: 45.2 versus 34.2 in the Noometry coding category.
Which has the bigger context window?
GPT-5.4 mini does, with 400K tokens against 262K.
How many benchmarks do GPT-5.4 mini and Mistral Medium share?
31 benchmarks have published results for both models. GPT-5.4 mini has 46 scored results on Noometry and Mistral Medium has 36.