Model comparison
GPT-5.4 mini vs Mistral Large 3
GPT-5.4 mini is the stronger model overall, scoring 45.0 to 39.1 on the Noometry Index. Mistral Large 3 costs 4.5× less per token, which makes it the better buy when GPT-5.4 mini's lead doesn't matter for your workload.
Last verified . 24 shared benchmarks.
Mistral AISummary
- They share 24 benchmarks with published results for both. GPT-5.4 mini scores higher in 7 categories and Mistral Large 3 in 2 categories; 6 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where GPT-5.4 mini leads 51.5 to 36.0.
- The biggest single-benchmark swing is NYT Connections (extended): 61.8% for GPT-5.4 mini and 7.5% for Mistral Large 3.
- Mistral Large 3 is cheaper at $0.25 / $0.75 per million input/output tokens, against $0.75 / $4.50 for GPT-5.4 mini.
- GPT-5.4 mini accepts more context: 400K tokens versus 262K.
- Mistral Large 3 has downloadable open weights; the other is API-only.
Side by side
| GPT-5.4 mini | Mistral Large 3 | |
|---|---|---|
| Provider | OpenAI | Mistral AI |
| Noometry Index | 45.0 | 39.1 |
| Released | 2026-03-17 | 2025-12-02 |
| Weights | Proprietary | Open |
| Context window | 400K | 262K |
| Max output | 128K | 8K |
| Input $ / M tokens | $0.75 | $0.25 |
| Output $ / M tokens | $4.50 | $0.75 |
| Results tracked | 46 | 24 |
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 Large 3: 34.4 (#237)
| Benchmark | GPT-5.4 mini | Mistral Large 3 |
|---|---|---|
| LMArena WebDev | 1397 | 1230 |
| LMArena Coding | 1438 | 1448 |
| FrontierCode | 27% | — |
| SciCode | 49.9% | — |
| WeirdML | 60.3% | — |
| ALE-Bench | 1,189 | — |
Agentic & Tool Use Not comparable
GPT-5.4 mini: 29.9 (#81), Mistral Large 3: —
| Benchmark | GPT-5.4 mini | Mistral Large 3 |
|---|---|---|
| DeepResearch Bench | 36.3% | — |
Reasoning GPT-5.4 mini leads
GPT-5.4 mini: 30.4 (#85), Mistral Large 3: 15.2 (#319)
| Benchmark | GPT-5.4 mini | Mistral Large 3 |
|---|---|---|
| Kagi LLM Benchmark | 37.9% | 50.9% |
| NYT Connections (extended) | 61.8% | 7.5% |
| Thematic Generalization | 61.7% | 23% |
| LMArena Hard Prompts | 1424 | 1429 |
| ARC-AGI-2 | 18.9% | — |
| ARC-AGI-1 | 63.7% | — |
| CritPt | 10% | — |
| Chess Puzzles | 24% | — |
| Mystery Game Puzzles | 11% | — |
| DTBench | 80% | — |
| LMCA | 40.8% | — |
| Epoch Capabilities Index | 148.84 | — |
| ForecastBench | 57 | — |
Math GPT-5.4 mini leads
GPT-5.4 mini: 45.5 (#75), Mistral Large 3: 38.7 (#129)
| Benchmark | GPT-5.4 mini | Mistral Large 3 |
|---|---|---|
| LMArena Math | 1419 | 1414 |
| FrontierMath (Tiers 1-3) | 51.2% | — |
| FrontierMath Tier 4 | 9.8% | — |
| OTIS Mock AIME 2024-2025 | 88.9% | — |
| ProofBench | 21% | — |
| FrontierMath (Feb 2025 set) | 28.3% | — |
| FrontierMath Tier 4 (v1) | 2.1% | — |
Knowledge GPT-5.4 mini leads
GPT-5.4 mini: 51.5 (#67), Mistral Large 3: 36.0 (#177)
| Benchmark | GPT-5.4 mini | Mistral Large 3 |
|---|---|---|
| Vectara Hallucination Rate | 5.5% | 14.5% |
| LMArena Expert | 1435 | 1421 |
| GPQA Diamond | 86.9% | — |
| SimpleQA Verified | 29.4% | — |
Multimodal GPT-5.4 mini leads
GPT-5.4 mini: 39.7 (#56), Mistral Large 3: 38.2 (#66)
| Benchmark | GPT-5.4 mini | Mistral Large 3 |
|---|---|---|
| LMArena Vision | 1245 | 1221 |
Multilingual Too close to call
GPT-5.4 mini: 51.9 (#96), Mistral Large 3: 52.5 (#84)
| Benchmark | GPT-5.4 mini | Mistral Large 3 |
|---|---|---|
| LMArena Non-English | 1405 | 1413 |
| LMArena Chinese | 1446 | 1447 |
| LMArena French | 1440 | 1455 |
| LMArena German | 1409 | 1437 |
| LMArena Japanese | 1374 | 1394 |
| LMArena Korean | 1368 | 1384 |
| LMArena Russian | 1417 | 1411 |
| LMArena Spanish | 1405 | 1440 |
Instruction Following Too close to call
GPT-5.4 mini: 74.1 (#102), Mistral Large 3: 74.0 (#108)
| Benchmark | GPT-5.4 mini | Mistral Large 3 |
|---|---|---|
| LMArena Instruction Following | 1405 | 1403 |
Long Context Too close to call
GPT-5.4 mini: 43.0 (#112), Mistral Large 3: 43.1 (#105)
| Benchmark | GPT-5.4 mini | Mistral Large 3 |
|---|---|---|
| LMArena Longer Query | 1407 | 1413 |
Writing & Preference GPT-5.4 mini leads
GPT-5.4 mini: 64.0 (#58), Mistral Large 3: 60.0 (#101)
| Benchmark | GPT-5.4 mini | Mistral Large 3 |
|---|---|---|
| LMArena Text | 1412 | 1428 |
| LMArena Creative Writing | 1370 | 1386 |
| EQ-Bench Creative Writing | 1665 | 1412 |
| LMArena Multi-Turn | 1429 | 1429 |
Frequently asked questions
Is GPT-5.4 mini better than Mistral Large 3?
GPT-5.4 mini is the stronger model overall, scoring 45.0 to 39.1 on the Noometry Index. Mistral Large 3 costs 4.5× less per token, which makes it the better buy when GPT-5.4 mini's lead doesn't matter for your workload.
Which is cheaper, GPT-5.4 mini or Mistral Large 3?
Mistral Large 3 is cheaper. It lists at $0.25 per million input tokens and $0.75 per million output tokens; GPT-5.4 mini lists at $0.75 and $4.50.
Is GPT-5.4 mini or Mistral Large 3 better for coding?
GPT-5.4 mini scores higher on coding benchmarks: 45.2 versus 34.4 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 Large 3 share?
24 benchmarks have published results for both models. GPT-5.4 mini has 46 scored results on Noometry and Mistral Large 3 has 24.