Model comparison
GPT-5 Nano vs MiniMax-M2.5
MiniMax-M2.5 is the stronger model overall, scoring 38.1 to 33.4 on the Noometry Index. GPT-5 Nano costs 3.8× less per token, which makes it the better buy when MiniMax-M2.5's lead doesn't matter for your workload.
Last verified . 25 shared benchmarks.
MiniMaxSummary
- They share 25 benchmarks with published results for both. GPT-5 Nano scores higher in 2 categories and MiniMax-M2.5 in 7 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where MiniMax-M2.5 leads 54.1 to 39.4.
- The biggest single-benchmark swing is ARC-AGI-1: 20.7% for GPT-5 Nano and 63.7% for MiniMax-M2.5.
- GPT-5 Nano is cheaper at $0.05 / $0.40 per million input/output tokens, against $0.30 / $1.20 for MiniMax-M2.5.
- GPT-5 Nano accepts more context: 400K tokens versus 205K.
- MiniMax-M2.5 has downloadable open weights; the other is API-only.
Side by side
| GPT-5 Nano | MiniMax-M2.5 | |
|---|---|---|
| Provider | OpenAI | MiniMax |
| Noometry Index | 33.4 | 38.1 |
| Released | 2025-08-07 | 2026-02-12 |
| Weights | Proprietary | Open |
| Context window | 400K | 205K |
| Max output | 128K | 131K |
| Input $ / M tokens | $0.05 | $0.30 |
| Output $ / M tokens | $0.40 | $1.20 |
| Results tracked | 48 | 33 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding MiniMax-M2.5 leads
GPT-5 Nano: 33.3 (#248), MiniMax-M2.5: 47.9 (#53)
| Benchmark | GPT-5 Nano | MiniMax-M2.5 |
|---|---|---|
| SWE-bench Verified (bash only) | 34.8% | 75.8% |
| LMArena Coding | 1351 | 1380 |
| ALE-Bench | 718.67 | 618.17 |
| LMArena WebDev | — | 1386 |
| SWE-bench Multilingual | — | 68.3% |
| WeirdML | 38.1% | — |
Agentic & Tool Use MiniMax-M2.5 leads
GPT-5 Nano: 25.6 (#99), MiniMax-M2.5: 30.2 (#70)
| Benchmark | GPT-5 Nano | MiniMax-M2.5 |
|---|---|---|
| Terminal-Bench | 21.8% | 42.7% |
| Berkeley Function Calling Leaderboard | 51.5% | — |
| Vending-Bench 2 | — | -23.16 |
Reasoning MiniMax-M2.5 leads
GPT-5 Nano: 16.4 (#291), MiniMax-M2.5: 17.6 (#281)
| Benchmark | GPT-5 Nano | MiniMax-M2.5 |
|---|---|---|
| ARC-AGI-2 | 2.6% | 4.9% |
| Kagi LLM Benchmark | 62.2% | 55.2% |
| ARC-AGI-1 | 20.7% | 63.7% |
| LMArena Hard Prompts | 1329 | 1372 |
| Epoch Capabilities Index | 139.39 | 146.68 |
| NYT Connections (extended) | — | 16.8% |
| Chess Puzzles | 27% | — |
| Mystery Game Puzzles | 9% | — |
| DTBench | 62.7% | — |
| LMCA | 7.9% | — |
| ForecastBench | 59.1 | — |
Math GPT-5 Nano leads
GPT-5 Nano: 29.1 (#221), MiniMax-M2.5: 25.8 (#242)
| Benchmark | GPT-5 Nano | MiniMax-M2.5 |
|---|---|---|
| ProofBench | 12% | 4% |
| LMArena Math | 1318 | 1378 |
| FrontierMath (Tiers 1-3) | 20% | — |
| FrontierMath Tier 4 | 2.4% | — |
| OTIS Mock AIME 2024-2025 | 81.1% | — |
| Omni-MATH | 54.6% | — |
| MATH Level 5 | 95.2% | — |
| FrontierMath (Feb 2025 set) | 8.3% | — |
| FrontierMath Tier 4 (v1) | 2.1% | — |
Knowledge MiniMax-M2.5 leads
GPT-5 Nano: 35.8 (#166), MiniMax-M2.5: 38.6 (#132)
| Benchmark | GPT-5 Nano | MiniMax-M2.5 |
|---|---|---|
| Vectara Hallucination Rate | 10.5% | 9.1% |
| LMArena Expert | 1321 | 1378 |
| GPQA Diamond | 69.4% | — |
| SimpleQA Verified | 11.7% | — |
| MMLU-Pro | 77.8% | — |
| GPQA (HELM) | 67.9% | — |
Multimodal Not comparable
GPT-5 Nano: 30.7 (#102), MiniMax-M2.5: —
| Benchmark | GPT-5 Nano | MiniMax-M2.5 |
|---|---|---|
| LMArena Vision | 1159 | — |
| VPCT | 37.2% | — |
Multilingual MiniMax-M2.5 leads
GPT-5 Nano: 45.5 (#167), MiniMax-M2.5: 47.3 (#147)
| Benchmark | GPT-5 Nano | MiniMax-M2.5 |
|---|---|---|
| LMArena Non-English | 1313 | 1338 |
| LMArena Chinese | 1355 | 1392 |
| LMArena German | 1329 | 1361 |
| LMArena Korean | 1269 | 1232 |
| LMArena Russian | 1296 | 1359 |
| LMArena Spanish | 1360 | 1354 |
| LMArena French | — | 1363 |
| LMArena Japanese | — | 1169 |
Instruction Following GPT-5 Nano leads
GPT-5 Nano: 75.2 (#72), MiniMax-M2.5: 71.7 (#143)
| Benchmark | GPT-5 Nano | MiniMax-M2.5 |
|---|---|---|
| LMArena Instruction Following | 1306 | 1353 |
| IFEval | 93.2% | — |
Long Context MiniMax-M2.5 leads
GPT-5 Nano: 31.3 (#275), MiniMax-M2.5: 37.4 (#211)
| Benchmark | GPT-5 Nano | MiniMax-M2.5 |
|---|---|---|
| LMArena Longer Query | 1313 | 1367 |
| Fiction.LiveBench | 44.4% | — |
| CL-bench | — | 11.4% |
| CL-bench Life | — | 6.3% |
Writing & Preference MiniMax-M2.5 leads
GPT-5 Nano: 39.4 (#244), MiniMax-M2.5: 54.1 (#149)
| Benchmark | GPT-5 Nano | MiniMax-M2.5 |
|---|---|---|
| LMArena Text | 1320 | 1359 |
| LMArena Creative Writing | 1250 | 1331 |
| EQ-Bench Creative Writing | 705 | 1361 |
| LMArena Multi-Turn | 1311 | 1364 |
| WildBench | 80.6% | — |
Frequently asked questions
Is GPT-5 Nano better than MiniMax-M2.5?
MiniMax-M2.5 is the stronger model overall, scoring 38.1 to 33.4 on the Noometry Index. GPT-5 Nano costs 3.8× less per token, which makes it the better buy when MiniMax-M2.5's lead doesn't matter for your workload.
Which is cheaper, GPT-5 Nano or MiniMax-M2.5?
GPT-5 Nano is cheaper. It lists at $0.05 per million input tokens and $0.40 per million output tokens; MiniMax-M2.5 lists at $0.30 and $1.20.
Is GPT-5 Nano or MiniMax-M2.5 better for coding?
MiniMax-M2.5 scores higher on coding benchmarks: 47.9 versus 33.3 in the Noometry coding category.
Which has the bigger context window?
GPT-5 Nano does, with 400K tokens against 205K.
How many benchmarks do GPT-5 Nano and MiniMax-M2.5 share?
25 benchmarks have published results for both models. GPT-5 Nano has 48 scored results on Noometry and MiniMax-M2.5 has 33.