Model comparison
GPT-5 Nano vs MiniMax-M3
MiniMax-M3 is the stronger model overall, scoring 43.5 to 33.4 on the Noometry Index. GPT-5 Nano costs 3.8× less per token, which makes it the better buy when MiniMax-M3's lead doesn't matter for your workload.
Last verified . 26 shared benchmarks.
MiniMaxSummary
- They share 26 benchmarks with published results for both. GPT-5 Nano scores higher in 1 category and MiniMax-M3 in 9 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where MiniMax-M3 leads 62.3 to 39.4.
- The biggest single-benchmark swing is LMCA: 7.9% for GPT-5 Nano and 33.7% for MiniMax-M3.
- GPT-5 Nano is cheaper at $0.05 / $0.40 per million input/output tokens, against $0.30 / $1.20 for MiniMax-M3.
- MiniMax-M3 accepts more context: 1M tokens versus 400K.
- MiniMax-M3 has downloadable open weights; the other is API-only.
Side by side
| GPT-5 Nano | MiniMax-M3 | |
|---|---|---|
| Provider | OpenAI | MiniMax |
| Noometry Index | 33.4 | 43.5 |
| Released | 2025-08-07 | 2026-06-01 |
| Weights | Proprietary | Open |
| Context window | 400K | 1M |
| Max output | 128K | 512K |
| Input $ / M tokens | $0.05 | $0.30 |
| Output $ / M tokens | $0.40 | $1.20 |
| Results tracked | 48 | 41 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding MiniMax-M3 leads
GPT-5 Nano: 33.3 (#248), MiniMax-M3: 41.5 (#111)
| Benchmark | GPT-5 Nano | MiniMax-M3 |
|---|---|---|
| LMArena Coding | 1351 | 1470 |
| ALE-Bench | 718.67 | 640.02 |
| FrontierCode | — | 14.7% |
| SWE-bench Verified (bash only) | 34.8% | — |
| LMArena WebDev | — | 1482 |
| SciCode | — | 47.1% |
| WeirdML | 38.1% | — |
Agentic & Tool Use GPT-5 Nano leads
GPT-5 Nano: 25.6 (#99), MiniMax-M3: 22.5 (#121)
| Benchmark | GPT-5 Nano | MiniMax-M3 |
|---|---|---|
| Terminal-Bench | 21.8% | — |
| APEX-Agents | — | 37.7% |
| Berkeley Function Calling Leaderboard | 51.5% | — |
| OSWorld 2.0 | — | 4.6% |
| GBAEval | — | 0.9% |
| Vending-Bench 2 | — | 2,158 |
Reasoning MiniMax-M3 leads
GPT-5 Nano: 16.4 (#291), MiniMax-M3: 29.3 (#83)
| Benchmark | GPT-5 Nano | MiniMax-M3 |
|---|---|---|
| Chess Puzzles | 27% | 14% |
| LMArena Hard Prompts | 1329 | 1447 |
| Mystery Game Puzzles | 9% | 8% |
| DTBench | 62.7% | 78.9% |
| LMCA | 7.9% | 33.7% |
| Epoch Capabilities Index | 139.39 | 147 |
| ForecastBench | 59.1 | 61.4 |
| ARC-AGI-2 | 2.6% | — |
| SimpleBench | — | 45.8% |
| Kagi LLM Benchmark | 62.2% | — |
| NYT Connections (extended) | — | 65.1% |
| ARC-AGI-1 | 20.7% | — |
| CritPt | — | 3.7% |
| Surface Evolver Bench | — | 55% |
Math MiniMax-M3 leads
GPT-5 Nano: 29.1 (#221), MiniMax-M3: 39.1 (#83)
| Benchmark | GPT-5 Nano | MiniMax-M3 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 81.1% | 71.1% |
| ProofBench | 12% | 18% |
| LMArena Math | 1318 | 1432 |
| FrontierMath (Tiers 1-3) | 20% | — |
| FrontierMath Tier 4 | 2.4% | — |
| Omni-MATH | 54.6% | — |
| MATH Level 5 | 95.2% | — |
| FrontierMath (Feb 2025 set) | 8.3% | — |
| FrontierMath Tier 4 (v1) | 2.1% | — |
Knowledge MiniMax-M3 leads
GPT-5 Nano: 35.8 (#166), MiniMax-M3: 58.4 (#32)
| Benchmark | GPT-5 Nano | MiniMax-M3 |
|---|---|---|
| GPQA Diamond | 69.4% | 90.9% |
| LMArena Expert | 1321 | 1462 |
| SimpleQA Verified | 11.7% | — |
| MMLU-Pro | 77.8% | — |
| Vectara Hallucination Rate | 10.5% | — |
| GPQA (HELM) | 67.9% | — |
Multimodal MiniMax-M3 leads
GPT-5 Nano: 30.7 (#102), MiniMax-M3: 39.5 (#45)
| Benchmark | GPT-5 Nano | MiniMax-M3 |
|---|---|---|
| LMArena Vision | 1159 | 1254 |
| VPCT | 37.2% | — |
| LMArena Document | — | 1435 |
Multilingual MiniMax-M3 leads
GPT-5 Nano: 45.5 (#167), MiniMax-M3: 53.2 (#70)
| Benchmark | GPT-5 Nano | MiniMax-M3 |
|---|---|---|
| LMArena Non-English | 1313 | 1420 |
| LMArena Chinese | 1355 | 1465 |
| LMArena German | 1329 | 1424 |
| LMArena Korean | 1269 | 1372 |
| LMArena Russian | 1296 | 1430 |
| LMArena Spanish | 1360 | 1431 |
| LMArena French | — | 1445 |
| LMArena Japanese | — | 1381 |
Instruction Following Too close to call
GPT-5 Nano: 75.2 (#72), MiniMax-M3: 75.7 (#56)
| Benchmark | GPT-5 Nano | MiniMax-M3 |
|---|---|---|
| LMArena Instruction Following | 1306 | 1433 |
| IFEval | 93.2% | — |
Long Context MiniMax-M3 leads
GPT-5 Nano: 31.3 (#275), MiniMax-M3: 44.2 (#68)
| Benchmark | GPT-5 Nano | MiniMax-M3 |
|---|---|---|
| LMArena Longer Query | 1313 | 1445 |
| Fiction.LiveBench | 44.4% | — |
Writing & Preference MiniMax-M3 leads
GPT-5 Nano: 39.4 (#244), MiniMax-M3: 62.3 (#79)
| Benchmark | GPT-5 Nano | MiniMax-M3 |
|---|---|---|
| LMArena Text | 1320 | 1433 |
| LMArena Creative Writing | 1250 | 1405 |
| LMArena Multi-Turn | 1311 | 1441 |
| EQ-Bench Creative Writing | 705 | — |
| WildBench | 80.6% | — |
| EQ-Bench 4 | — | 1150 |
Frequently asked questions
Is GPT-5 Nano better than MiniMax-M3?
MiniMax-M3 is the stronger model overall, scoring 43.5 to 33.4 on the Noometry Index. GPT-5 Nano costs 3.8× less per token, which makes it the better buy when MiniMax-M3's lead doesn't matter for your workload.
Which is cheaper, GPT-5 Nano or MiniMax-M3?
GPT-5 Nano is cheaper. It lists at $0.05 per million input tokens and $0.40 per million output tokens; MiniMax-M3 lists at $0.30 and $1.20.
Is GPT-5 Nano or MiniMax-M3 better for coding?
MiniMax-M3 scores higher on coding benchmarks: 41.5 versus 33.3 in the Noometry coding category.
Which has the bigger context window?
MiniMax-M3 does, with 1M tokens against 400K.
How many benchmarks do GPT-5 Nano and MiniMax-M3 share?
26 benchmarks have published results for both models. GPT-5 Nano has 48 scored results on Noometry and MiniMax-M3 has 41.