Model comparison
GPT-5 Mini vs MiniMax-M2.5
GPT-5 Mini is the stronger model overall, scoring 41.8 to 38.3 on the Noometry Index.
Last verified . 29 shared benchmarks.
MiniMaxSummary
- They share 29 benchmarks with published results for both. GPT-5 Mini scores higher in 8 categories and MiniMax-M2.5 in 1 category; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where GPT-5 Mini leads 46.7 to 26.9.
- The biggest single-benchmark swing is SWE-bench Multilingual: 39.7% for GPT-5 Mini and 68.3% for MiniMax-M2.5.
- MiniMax-M2.5 is cheaper at $0.30 / $1.20 per million input/output tokens, against $0.25 / $2 for GPT-5 Mini.
- GPT-5 Mini accepts more context: 400K tokens versus 205K.
- MiniMax-M2.5 has downloadable open weights; the other is API-only.
Side by side
| GPT-5 Mini | MiniMax-M2.5 | |
|---|---|---|
| Provider | OpenAI | MiniMax |
| Noometry Index | 41.8 | 38.3 |
| Released | 2025-08-07 | 2026-02-12 |
| Weights | Proprietary | Open |
| Context window | 400K | 205K |
| Max output | 128K | 131K |
| Input $ / M tokens | $0.25 | $0.30 |
| Output $ / M tokens | $2 | $1.20 |
| Results tracked | 60 | 33 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding MiniMax-M2.5 leads
GPT-5 Mini: 40.1 (#146), MiniMax-M2.5: 48.1 (#58)
| Benchmark | GPT-5 Mini | MiniMax-M2.5 |
|---|---|---|
| SWE-bench Verified (bash only) | 59.8% | 75.8% |
| SWE-bench Multilingual | 39.7% | 68.3% |
| LMArena Coding | 1406 | 1381 |
| ALE-Bench | 799.77 | 618.17 |
| SWE-bench Verified | 64.7% | — |
| LMArena WebDev | — | 1387 |
| SciCode | 39.2% | — |
| WeirdML | 52.7% | — |
| AlgoTune | 1.38 | — |
Agentic & Tool Use Too close to call
GPT-5 Mini: 31.1 (#70), MiniMax-M2.5: 30.4 (#77)
| Benchmark | GPT-5 Mini | MiniMax-M2.5 |
|---|---|---|
| Terminal-Bench | 34.8% | 42.7% |
| Vending-Bench 2 | -31.18 | -23.16 |
| Berkeley Function Calling Leaderboard | 55.5% | — |
Reasoning GPT-5 Mini leads
GPT-5 Mini: 23.9 (#168), MiniMax-M2.5: 17.5 (#292)
| Benchmark | GPT-5 Mini | MiniMax-M2.5 |
|---|---|---|
| ARC-AGI-2 | 4.4% | 4.9% |
| Kagi LLM Benchmark | 70.3% | 55.2% |
| ARC-AGI-1 | 54.3% | 63.7% |
| LMArena Hard Prompts | 1380 | 1372 |
| Epoch Capabilities Index | 145.52 | 146.68 |
| NYT Connections (extended) | — | 16.8% |
| CritPt | 0% | — |
| Chess Puzzles | 30% | — |
| EnigmaEval | 8.2% | — |
| Mystery Game Puzzles | 10% | — |
| DTBench | 80.5% | — |
| LMCA | 34.2% | — |
| ForecastBench | 61 | — |
Math GPT-5 Mini leads
GPT-5 Mini: 46.7 (#69), MiniMax-M2.5: 26.9 (#253)
| Benchmark | GPT-5 Mini | MiniMax-M2.5 |
|---|---|---|
| ProofBench | 9% | 4% |
| LMArena Math | 1378 | 1378 |
| FrontierMath (Tiers 1-3) | 46.7% | — |
| FrontierMath Tier 4 | 12.2% | — |
| OTIS Mock AIME 2024-2025 | 86.7% | — |
| Omni-MATH | 72.2% | — |
| MATH Level 5 | 97.8% | — |
| FrontierMath (Feb 2025 set) | 27.2% | — |
| FrontierMath Tier 4 (v1) | 6.3% | — |
Knowledge GPT-5 Mini leads
GPT-5 Mini: 45.6 (#86), MiniMax-M2.5: 39.2 (#135)
| Benchmark | GPT-5 Mini | MiniMax-M2.5 |
|---|---|---|
| Vectara Hallucination Rate | 12.9% | 9.1% |
| LMArena Expert | 1379 | 1379 |
| GPQA Diamond | 75% | — |
| Humanity's Last Exam | 19.4% | — |
| SimpleQA Verified | 21.6% | — |
| MMLU-Pro | 83.5% | — |
| Confabulations | 13.3% | — |
| GPQA (HELM) | 75.6% | — |
Multimodal Not comparable
GPT-5 Mini: 35.6 (#85), MiniMax-M2.5: —
| Benchmark | GPT-5 Mini | MiniMax-M2.5 |
|---|---|---|
| LMArena Vision | 1202 | — |
| VPCT | 40.2% | — |
Multilingual GPT-5 Mini leads
GPT-5 Mini: 48.9 (#137), MiniMax-M2.5: 47.1 (#152)
| Benchmark | GPT-5 Mini | MiniMax-M2.5 |
|---|---|---|
| LMArena Non-English | 1363 | 1338 |
| LMArena Chinese | 1385 | 1393 |
| LMArena French | 1386 | 1362 |
| LMArena German | 1366 | 1362 |
| LMArena Japanese | 1341 | 1171 |
| LMArena Korean | 1308 | 1232 |
| LMArena Russian | 1362 | 1358 |
| LMArena Spanish | 1355 | 1354 |
Instruction Following GPT-5 Mini leads
GPT-5 Mini: 76.2 (#46), MiniMax-M2.5: 71.5 (#148)
| Benchmark | GPT-5 Mini | MiniMax-M2.5 |
|---|---|---|
| LMArena Instruction Following | 1357 | 1353 |
| IFEval | 92.7% | — |
Long Context GPT-5 Mini leads
GPT-5 Mini: 41.9 (#132), MiniMax-M2.5: 37.5 (#216)
| Benchmark | GPT-5 Mini | MiniMax-M2.5 |
|---|---|---|
| LMArena Longer Query | 1355 | 1366 |
| Fiction.LiveBench | 69.4% | — |
| CL-bench | — | 11.4% |
| CL-bench Life | — | 6.3% |
Writing & Preference GPT-5 Mini leads
GPT-5 Mini: 55.2 (#148), MiniMax-M2.5: 53.9 (#153)
| Benchmark | GPT-5 Mini | MiniMax-M2.5 |
|---|---|---|
| LMArena Text | 1373 | 1359 |
| LMArena Creative Writing | 1325 | 1331 |
| EQ-Bench Creative Writing | 1313 | 1361 |
| LMArena Multi-Turn | 1363 | 1364 |
| Short-Story Creative Writing | 83.1% | — |
| WildBench | 85.5% | — |
Frequently asked questions
Is GPT-5 Mini better than MiniMax-M2.5?
GPT-5 Mini is the stronger model overall, scoring 41.8 to 38.3 on the Noometry Index.
Which is cheaper, GPT-5 Mini or MiniMax-M2.5?
MiniMax-M2.5 is cheaper. It lists at $0.30 per million input tokens and $1.20 per million output tokens; GPT-5 Mini lists at $0.25 and $2.
Is GPT-5 Mini or MiniMax-M2.5 better for coding?
MiniMax-M2.5 scores higher on coding benchmarks: 48.1 versus 40.1 in the Noometry coding category.
Which has the bigger context window?
GPT-5 Mini does, with 400K tokens against 205K.
How many benchmarks do GPT-5 Mini and MiniMax-M2.5 share?
29 benchmarks have published results for both models. GPT-5 Mini has 60 scored results on Noometry and MiniMax-M2.5 has 33.