Model comparison
GPT-5 Nano vs Magistral Medium
Magistral Medium is the stronger model overall, scoring 34.8 to 33.4 on the Noometry Index. GPT-5 Nano costs 20× less per token, which makes it the better buy when Magistral Medium's lead doesn't matter for your workload.
Last verified . 18 shared benchmarks.
Mistral AISummary
- They share 18 benchmarks with published results for both. GPT-5 Nano scores higher in 4 categories and Magistral Medium in 4 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in instruction following, where GPT-5 Nano leads 75.2 to 66.2.
- The biggest single-benchmark swing is Kagi LLM Benchmark: 62.2% for GPT-5 Nano and 16.2% for Magistral Medium.
- GPT-5 Nano is cheaper at $0.05 / $0.40 per million input/output tokens, against $2 / $5 for Magistral Medium.
- GPT-5 Nano accepts more context: 400K tokens versus 128K.
- Magistral Medium has downloadable open weights; the other is API-only.
Side by side
| GPT-5 Nano | Magistral Medium | |
|---|---|---|
| Provider | OpenAI | Mistral AI |
| Noometry Index | 33.4 | 34.8 |
| Released | 2025-08-07 | 2025-03-17 |
| Weights | Proprietary | Open |
| Context window | 400K | 128K |
| Max output | 128K | 16K |
| Input $ / M tokens | $0.05 | $2 |
| Output $ / M tokens | $0.40 | $5 |
| Results tracked | 48 | 22 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Magistral Medium leads
GPT-5 Nano: 33.3 (#248), Magistral Medium: 38.6 (#161)
| Benchmark | GPT-5 Nano | Magistral Medium |
|---|---|---|
| LMArena Coding | 1351 | 1319 |
| SWE-bench Verified (bash only) | 34.8% | — |
| SciCode | — | 39.2% |
| WeirdML | 38.1% | — |
| ALE-Bench | 718.67 | — |
Agentic & Tool Use Not comparable
GPT-5 Nano: 25.6 (#99), Magistral Medium: —
| Benchmark | GPT-5 Nano | Magistral Medium |
|---|---|---|
| Terminal-Bench | 21.8% | — |
| Berkeley Function Calling Leaderboard | 51.5% | — |
Reasoning GPT-5 Nano leads
GPT-5 Nano: 16.4 (#291), Magistral Medium: 8.7 (#342)
| Benchmark | GPT-5 Nano | Magistral Medium |
|---|---|---|
| ARC-AGI-2 | 2.6% | 0% |
| Kagi LLM Benchmark | 62.2% | 16.2% |
| ARC-AGI-1 | 20.7% | 6.1% |
| LMArena Hard Prompts | 1329 | 1267 |
| CritPt | — | 0.3% |
| Chess Puzzles | 27% | — |
| Mystery Game Puzzles | 9% | — |
| DTBench | 62.7% | — |
| LMCA | 7.9% | — |
| Epoch Capabilities Index | 139.39 | — |
| ForecastBench | 59.1 | — |
Math Magistral Medium leads
GPT-5 Nano: 29.1 (#221), Magistral Medium: 32.4 (#187)
| Benchmark | GPT-5 Nano | Magistral Medium |
|---|---|---|
| LMArena Math | 1318 | 1249 |
| FrontierMath (Tiers 1-3) | 20% | — |
| FrontierMath Tier 4 | 2.4% | — |
| OTIS Mock AIME 2024-2025 | 81.1% | — |
| ProofBench | 12% | — |
| Omni-MATH | 54.6% | — |
| MATH Level 5 | 95.2% | — |
| FrontierMath (Feb 2025 set) | 8.3% | — |
| FrontierMath Tier 4 (v1) | 2.1% | — |
Knowledge GPT-5 Nano leads
GPT-5 Nano: 35.8 (#166), Magistral Medium: 32.9 (#199)
| Benchmark | GPT-5 Nano | Magistral Medium |
|---|---|---|
| LMArena Expert | 1321 | 1222 |
| GPQA Diamond | 69.4% | — |
| SimpleQA Verified | 11.7% | — |
| MMLU-Pro | 77.8% | — |
| Vectara Hallucination Rate | 10.5% | — |
| GPQA (HELM) | 67.9% | — |
Multimodal Not comparable
GPT-5 Nano: 30.7 (#102), Magistral Medium: —
| Benchmark | GPT-5 Nano | Magistral Medium |
|---|---|---|
| LMArena Vision | 1159 | — |
| VPCT | 37.2% | — |
Multilingual GPT-5 Nano leads
GPT-5 Nano: 45.5 (#167), Magistral Medium: 39.8 (#219)
| Benchmark | GPT-5 Nano | Magistral Medium |
|---|---|---|
| LMArena Non-English | 1313 | 1232 |
| LMArena Chinese | 1355 | 1225 |
| LMArena German | 1329 | 1248 |
| LMArena Korean | 1269 | 1125 |
| LMArena Russian | 1296 | 1224 |
| LMArena Spanish | 1360 | 1269 |
| LMArena French | — | 1268 |
| LMArena Japanese | — | 1176 |
Instruction Following GPT-5 Nano leads
GPT-5 Nano: 75.2 (#72), Magistral Medium: 66.2 (#207)
| Benchmark | GPT-5 Nano | Magistral Medium |
|---|---|---|
| LMArena Instruction Following | 1306 | 1254 |
| IFEval | 93.2% | — |
Long Context Magistral Medium leads
GPT-5 Nano: 31.3 (#275), Magistral Medium: 39.3 (#178)
| Benchmark | GPT-5 Nano | Magistral Medium |
|---|---|---|
| LMArena Longer Query | 1313 | 1295 |
| Fiction.LiveBench | 44.4% | — |
Writing & Preference Magistral Medium leads
GPT-5 Nano: 39.4 (#244), Magistral Medium: 46.5 (#214)
| Benchmark | GPT-5 Nano | Magistral Medium |
|---|---|---|
| LMArena Text | 1320 | 1255 |
| LMArena Creative Writing | 1250 | 1245 |
| LMArena Multi-Turn | 1311 | 1275 |
| EQ-Bench Creative Writing | 705 | — |
| WildBench | 80.6% | — |
Frequently asked questions
Is GPT-5 Nano better than Magistral Medium?
Magistral Medium is the stronger model overall, scoring 34.8 to 33.4 on the Noometry Index. GPT-5 Nano costs 20× less per token, which makes it the better buy when Magistral Medium's lead doesn't matter for your workload.
Which is cheaper, GPT-5 Nano or Magistral Medium?
GPT-5 Nano is cheaper. It lists at $0.05 per million input tokens and $0.40 per million output tokens; Magistral Medium lists at $2 and $5.
Is GPT-5 Nano or Magistral Medium better for coding?
Magistral Medium scores higher on coding benchmarks: 38.6 versus 33.3 in the Noometry coding category.
Which has the bigger context window?
GPT-5 Nano does, with 400K tokens against 128K.
How many benchmarks do GPT-5 Nano and Magistral Medium share?
18 benchmarks have published results for both models. GPT-5 Nano has 48 scored results on Noometry and Magistral Medium has 22.