# GPT-5.4 nano vs Mistral Small

> GPT-5.4 nano is the stronger model overall, scoring 41.9 to 33.1 on the Noometry Index. Mistral Small costs 1.8× less per token, which makes it the better buy when GPT-5.4 nano's lead doesn't matter for your workload.

- Canonical page: https://noometry.com/compare/gpt-5-4-nano-vs-mistral-small
- Last updated: 2026-10-10
- Shared benchmarks: 27

## Summary

- They share 27 benchmarks with published results for both. GPT-5.4 nano scores higher in 9 categories and Mistral Small in 0 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in math, where GPT-5.4 nano leads 40.7 to 15.4.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 87.8% for GPT-5.4 nano and 5.8% for Mistral Small.
- Mistral Small is cheaper at $0.15 / $0.60 per million input/output tokens, against $0.20 / $1.25 for GPT-5.4 nano.
- GPT-5.4 nano accepts more context: 400K tokens versus 256K.
- Mistral Small has downloadable open weights; the other is API-only.

## Snapshot

| | GPT-5.4 nano | Mistral Small |
|---|---|---|
| Provider | OpenAI | Mistral AI |
| Noometry Index | 41.9 | 33.1 |
| Rank | 111 | 238 |
| Context | 400K | 256K |
| Input $/M | $0.20 | $0.15 |
| Output $/M | $1.25 | $0.60 |
| Weights | Proprietary | Open |

## Coding

- GPT-5.4 nano: 43.1 (#82)
- Mistral Small: 33.5 (#242)

| Benchmark | GPT-5.4 nano | Mistral Small |
|---|---|---|
| SciCode | 46.9% | 26.5% |
| LMArena Coding | 1404 | 1363 |
| ALE-Bench | 1,005 | 497.62 |
| WeirdML | 49.2% | — |
| BigCodeBench Instruct | — | 36.1% |
| LiveBench Coding | — | 36.2% |
| BigCodeBench Complete | — | 46.6% |

## Agentic & Tool Use

- GPT-5.4 nano: —
- Mistral Small: 27.9 (#87)

| Benchmark | GPT-5.4 nano | Mistral Small |
|---|---|---|
| Berkeley Function Calling Leaderboard | — | 37.1% |

## Reasoning

- GPT-5.4 nano: 23.8 (#159)
- Mistral Small: 19.3 (#238)

| Benchmark | GPT-5.4 nano | Mistral Small |
|---|---|---|
| Kagi LLM Benchmark | 39.7% | 37.8% |
| CritPt | 9.3% | 0% |
| LMArena Hard Prompts | 1380 | 1336 |
| DTBench | 80.3% | 70.9% |
| LMCA | 36.9% | 20.6% |
| ARC-AGI-2 | 5.7% | — |
| ARC-AGI-1 | 51.5% | — |
| Chess Puzzles | 30% | — |
| LiveBench Reasoning | — | 44.8% |
| Mystery Game Puzzles | 9% | — |
| LiveBench Data Analysis | — | 53.7% |
| Epoch Capabilities Index | 145.78 | — |
| ForecastBench | 57.3 | — |
| LiveBench | — | 44% |

## Math

- GPT-5.4 nano: 40.7 (#76)
- Mistral Small: 15.4 (#287)

| Benchmark | GPT-5.4 nano | Mistral Small |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 87.8% | 5.8% |
| LMArena Math | 1406 | 1341 |
| FrontierMath (Tiers 1-3) | 44.9% | — |
| FrontierMath Tier 4 | 12.2% | — |
| ProofBench | 5% | — |
| LiveBench Math | — | 39.9% |
| MATH Level 5 | — | 46.8% |
| FrontierMath (Feb 2025 set) | 25.9% | — |
| FrontierMath Tier 4 (v1) | 6.3% | — |

## Knowledge

- GPT-5.4 nano: 41.8 (#94)
- Mistral Small: 30.9 (#214)

| Benchmark | GPT-5.4 nano | Mistral Small |
|---|---|---|
| GPQA Diamond | 78.5% | 47.5% |
| Vectara Hallucination Rate | 3.1% | 5.1% |
| LMArena Expert | 1397 | 1293 |
| SimpleQA Verified | 11.7% | — |
| MMLU | — | 68.7% |

## Multimodal

- GPT-5.4 nano: 36.0 (#73)
- Mistral Small: 32.9 (#91)

| Benchmark | GPT-5.4 nano | Mistral Small |
|---|---|---|
| LMArena Vision | 1196 | 1142 |

## Multilingual

- GPT-5.4 nano: 48.8 (#136)
- Mistral Small: 45.7 (#164)

| Benchmark | GPT-5.4 nano | Mistral Small |
|---|---|---|
| LMArena Non-English | 1359 | 1315 |
| LMArena Chinese | 1391 | 1342 |
| LMArena French | 1396 | 1337 |
| LMArena German | 1368 | 1341 |
| LMArena Japanese | 1342 | 1275 |
| LMArena Korean | 1321 | 1260 |
| LMArena Russian | 1363 | 1324 |
| LMArena Spanish | 1368 | 1346 |

## Instruction Following

- GPT-5.4 nano: 72.1 (#139)
- Mistral Small: 66.6 (#204)

| Benchmark | GPT-5.4 nano | Mistral Small |
|---|---|---|
| LMArena Instruction Following | 1362 | 1310 |
| LiveBench Instruction Following | — | 63.7% |

## Long Context

- GPT-5.4 nano: 41.6 (#131)
- Mistral Small: 40.4 (#151)

| Benchmark | GPT-5.4 nano | Mistral Small |
|---|---|---|
| LMArena Longer Query | 1365 | 1327 |

## Writing & Preference

- GPT-5.4 nano: 55.9 (#137)
- Mistral Small: 52.6 (#165)

| Benchmark | GPT-5.4 nano | Mistral Small |
|---|---|---|
| LMArena Text | 1372 | 1338 |
| LMArena Creative Writing | 1313 | 1304 |
| LMArena Multi-Turn | 1382 | 1344 |
| LiveBench Language | — | 30.5% |

## FAQ

### Is GPT-5.4 nano better than Mistral Small?

GPT-5.4 nano is the stronger model overall, scoring 41.9 to 33.1 on the Noometry Index. Mistral Small costs 1.8× less per token, which makes it the better buy when GPT-5.4 nano's lead doesn't matter for your workload.

### Which is cheaper, GPT-5.4 nano or Mistral Small?

Mistral Small is cheaper. It lists at $0.15 per million input tokens and $0.60 per million output tokens; GPT-5.4 nano lists at $0.20 and $1.25.

### Is GPT-5.4 nano or Mistral Small better for coding?

GPT-5.4 nano scores higher on coding benchmarks: 43.1 versus 33.5 in the Noometry coding category.

### Which has the bigger context window?

GPT-5.4 nano does, with 400K tokens against 256K.

### How many benchmarks do GPT-5.4 nano and Mistral Small share?

27 benchmarks have published results for both models. GPT-5.4 nano has 40 scored results on Noometry and Mistral Small has 39.