# GPT-5.5 vs Mistral Medium

> GPT-5.5 is the stronger model overall, scoring 63.4 to 36.0 on the Noometry Index. Mistral Medium costs 3.8× less per token, which makes it the better buy when GPT-5.5's lead doesn't matter for your workload.

- Canonical page: https://noometry.com/compare/gpt-5-5-vs-mistral-medium
- Last updated: 2026-10-10
- Shared benchmarks: 32

## Summary

- They share 32 benchmarks with published results for both. GPT-5.5 scores higher in 10 categories and Mistral Medium in 0 categories; 10 gaps are clear of the uncertainty.
- The widest gap is in math, where GPT-5.5 leads 81.5 to 27.2.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 100% for GPT-5.5 and 32.2% for Mistral Medium.
- Mistral Medium is cheaper at $1.50 / $7.50 per million input/output tokens, against $5 / $30 for GPT-5.5.
- GPT-5.5 accepts more context: 1.05M tokens versus 262K.
- Mistral Medium has downloadable open weights; the other is API-only.

## Snapshot

| | GPT-5.5 | Mistral Medium |
|---|---|---|
| Provider | OpenAI | Mistral AI |
| Noometry Index | 63.4 | 36.0 |
| Rank | 8 | 211 |
| Context | 1.05M | 262K |
| Input $/M | $5 | $1.50 |
| Output $/M | $30 | $7.50 |
| Weights | Proprietary | Open |

## Coding

- GPT-5.5: 58.3 (#13)
- Mistral Medium: 33.8 (#241)

| Benchmark | GPT-5.5 | Mistral Medium |
|---|---|---|
| FrontierCode | 43% | 8% |
| SciCode | 56.1% | 40.2% |
| WeirdML | 84.9% | 43.7% |
| LMArena Coding | 1494 | 1434 |
| ALE-Bench | 1,943 | 763.98 |
| SWE-bench Verified | 80.6% | — |
| DeepSWE | 67% | — |
| LMArena WebDev | 1512 | — |
| GSO | 40.2% | — |
| MirrorCode | 10% | — |

## Agentic & Tool Use

- GPT-5.5: 50.7 (#6)
- Mistral Medium: 28.1 (#84)

| Benchmark | GPT-5.5 | Mistral Medium |
|---|---|---|
| Terminal-Bench | 84.7% | — |
| APEX-Agents | 55.1% | — |
| Berkeley Function Calling Leaderboard | — | 37.7% |
| OSWorld 2.0 | 13% | — |
| Remote Labor Index | 6.3% | — |
| τ²-bench Banking | 44.6% | — |
| DeepResearch Bench | 54% | — |
| PostTrainBench | 27.2% | — |
| ExploitBench | 47.4% | — |
| GBAEval | 53.2% | — |
| GDP.pdf | 26% | — |
| LMArena Search | 1242 | — |
| Vending-Bench 2 | 7,524 | — |

## Reasoning

- GPT-5.5: 72.7 (#8)
- Mistral Medium: 23.3 (#165)

| Benchmark | GPT-5.5 | Mistral Medium |
|---|---|---|
| Kagi LLM Benchmark | 88.8% | 50% |
| CritPt | 27.1% | 0% |
| LMArena Hard Prompts | 1489 | 1426 |
| DTBench | 96% | 75.5% |
| LMCA | 54.3% | 26.1% |
| Surface Evolver Bench | 88.1% | 26.9% |
| ARC-AGI-2 | 85% | — |
| SimpleBench | 69% | — |
| NYT Connections (extended) | 96.2% | — |
| ARC-AGI-1 | 95% | — |
| Chess Puzzles | 54% | — |
| EBR-Bench | 34.3% | — |
| Mystery Game Puzzles | 56% | — |
| Bench to the Future 3 | 0.14 | — |
| Epoch Capabilities Index | 159.25 | — |
| ForecastBench | 60.6 | — |

## Math

- GPT-5.5: 81.5 (#9)
- Mistral Medium: 27.2 (#238)

| Benchmark | GPT-5.5 | Mistral Medium |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 100% | 32.2% |
| ProofBench | 50% | 9% |
| LMArena Math | 1486 | 1409 |
| FrontierMath (Feb 2025 set) | 51.7% | 0.3% |
| FrontierMath (Tiers 1-3) | 85.3% | — |
| FrontierMath Tier 4 | 72.5% | — |
| MathArena Final-Answer Competitions | 94.3% | — |
| MATH Level 5 | — | 81.6% |
| FrontierMath Erdős | 0% | — |
| FrontierMath Tier 4 (v1) | 35.4% | — |

## Knowledge

- GPT-5.5: 64.4 (#14)
- Mistral Medium: 24.9 (#260)

| Benchmark | GPT-5.5 | Mistral Medium |
|---|---|---|
| GPQA Diamond | 94% | 59.5% |
| Vectara Hallucination Rate | 9.3% | 22.7% |
| LMArena Expert | 1508 | 1408 |
| Humanity's Last Exam | — | 4.5% |
| SimpleQA Verified | 63% | — |

## Multimodal

- GPT-5.5: 47.1 (#11)
- Mistral Medium: 34.6 (#82)

| Benchmark | GPT-5.5 | Mistral Medium |
|---|---|---|
| LMArena Vision | 1297 | 1172 |
| Blueprint-Bench 2 | 36.2% | — |
| Furniture Assembly | 44.2% | — |
| LMArena Document | 1486 | — |

## Multilingual

- GPT-5.5: 56.5 (#19)
- Mistral Medium: 52.3 (#86)

| Benchmark | GPT-5.5 | Mistral Medium |
|---|---|---|
| LMArena Non-English | 1467 | 1408 |
| LMArena Chinese | 1535 | 1448 |
| LMArena French | 1486 | 1460 |
| LMArena German | 1478 | 1433 |
| LMArena Japanese | 1498 | 1378 |
| LMArena Korean | 1461 | 1382 |
| LMArena Russian | 1473 | 1411 |
| LMArena Spanish | 1469 | 1433 |

## Instruction Following

- GPT-5.5: 77.7 (#18)
- Mistral Medium: 73.9 (#111)

| Benchmark | GPT-5.5 | Mistral Medium |
|---|---|---|
| LMArena Instruction Following | 1479 | 1398 |

## Long Context

- GPT-5.5: 48.3 (#12)
- Mistral Medium: 42.9 (#108)

| Benchmark | GPT-5.5 | Mistral Medium |
|---|---|---|
| LMArena Longer Query | 1484 | 1406 |
| CL-bench Life | 22.2% | — |

## Writing & Preference

- GPT-5.5: 72.8 (#13)
- Mistral Medium: 60.2 (#97)

| Benchmark | GPT-5.5 | Mistral Medium |
|---|---|---|
| LMArena Text | 1471 | 1424 |
| LMArena Creative Writing | 1455 | 1391 |
| LMArena Multi-Turn | 1477 | 1418 |
| Short-Story Creative Writing | — | 77.3% |
| EQ-Bench Creative Writing | 1844 | — |
| EQ-Bench 4 | 1315 | — |

## FAQ

### Is GPT-5.5 better than Mistral Medium?

GPT-5.5 is the stronger model overall, scoring 63.4 to 36.0 on the Noometry Index. Mistral Medium costs 3.8× less per token, which makes it the better buy when GPT-5.5's lead doesn't matter for your workload.

### Which is cheaper, GPT-5.5 or Mistral Medium?

Mistral Medium is cheaper. It lists at $1.50 per million input tokens and $7.50 per million output tokens; GPT-5.5 lists at $5 and $30.

### Is GPT-5.5 or Mistral Medium better for coding?

GPT-5.5 scores higher on coding benchmarks: 58.3 versus 33.8 in the Noometry coding category.

### Which has the bigger context window?

GPT-5.5 does, with 1.05M tokens against 262K.

### How many benchmarks do GPT-5.5 and Mistral Medium share?

32 benchmarks have published results for both models. GPT-5.5 has 71 scored results on Noometry and Mistral Medium has 36.