Model comparison
GPT-5.4 mini vs MiniMax-M2.5
GPT-5.4 mini is the stronger model overall, scoring 44.8 to 38.1 on the Noometry Index. MiniMax-M2.5 costs 3.2× less per token, which makes it the better buy when GPT-5.4 mini's lead doesn't matter for your workload.
Last verified . 27 shared benchmarks.
MiniMaxSummary
- They share 27 benchmarks with published results for both. GPT-5.4 mini scores higher in 7 categories and MiniMax-M2.5 in 2 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where GPT-5.4 mini leads 45.4 to 25.8.
- The biggest single-benchmark swing is NYT Connections (extended): 61.8% for GPT-5.4 mini and 16.8% for MiniMax-M2.5.
- MiniMax-M2.5 is cheaper at $0.30 / $1.20 per million input/output tokens, against $0.75 / $4.50 for GPT-5.4 mini.
- GPT-5.4 mini accepts more context: 400K tokens versus 205K.
- MiniMax-M2.5 has downloadable open weights; the other is API-only.
Side by side
| GPT-5.4 mini | MiniMax-M2.5 | |
|---|---|---|
| Provider | OpenAI | MiniMax |
| Noometry Index | 44.8 | 38.1 |
| Released | 2026-03-17 | 2026-02-12 |
| Weights | Proprietary | Open |
| Context window | 400K | 205K |
| Max output | 128K | 131K |
| Input $ / M tokens | $0.75 | $0.30 |
| Output $ / M tokens | $4.50 | $1.20 |
| Results tracked | 46 | 33 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding MiniMax-M2.5 leads
GPT-5.4 mini: 45.0 (#67), MiniMax-M2.5: 47.9 (#53)
| Benchmark | GPT-5.4 mini | MiniMax-M2.5 |
|---|---|---|
| LMArena WebDev | 1397 | 1386 |
| LMArena Coding | 1438 | 1380 |
| ALE-Bench | 1,189 | 618.17 |
| FrontierCode | 27% | — |
| SWE-bench Verified (bash only) | — | 75.8% |
| SWE-bench Multilingual | — | 68.3% |
| SciCode | 49.9% | — |
| WeirdML | 60.3% | — |
Agentic & Tool Use Too close to call
GPT-5.4 mini: 29.6 (#74), MiniMax-M2.5: 30.2 (#70)
| Benchmark | GPT-5.4 mini | MiniMax-M2.5 |
|---|---|---|
| Terminal-Bench | — | 42.7% |
| DeepResearch Bench | 36.3% | — |
| Vending-Bench 2 | — | -23.16 |
Reasoning GPT-5.4 mini leads
GPT-5.4 mini: 30.3 (#77), MiniMax-M2.5: 17.6 (#281)
| Benchmark | GPT-5.4 mini | MiniMax-M2.5 |
|---|---|---|
| ARC-AGI-2 | 18.9% | 4.9% |
| Kagi LLM Benchmark | 37.9% | 55.2% |
| NYT Connections (extended) | 61.8% | 16.8% |
| ARC-AGI-1 | 63.7% | 63.7% |
| LMArena Hard Prompts | 1424 | 1372 |
| Epoch Capabilities Index | 148.83 | 146.68 |
| CritPt | 10% | — |
| Chess Puzzles | 24% | — |
| Thematic Generalization | 61.7% | — |
| Mystery Game Puzzles | 11% | — |
| DTBench | 80% | — |
| LMCA | 40.8% | — |
| ForecastBench | 57 | — |
Math GPT-5.4 mini leads
GPT-5.4 mini: 45.4 (#63), MiniMax-M2.5: 25.8 (#242)
| Benchmark | GPT-5.4 mini | MiniMax-M2.5 |
|---|---|---|
| ProofBench | 21% | 4% |
| LMArena Math | 1418 | 1378 |
| FrontierMath (Tiers 1-3) | 51.2% | — |
| FrontierMath Tier 4 | 9.8% | — |
| OTIS Mock AIME 2024-2025 | 88.9% | — |
| FrontierMath (Feb 2025 set) | 28.3% | — |
| FrontierMath Tier 4 (v1) | 2.1% | — |
Knowledge GPT-5.4 mini leads
GPT-5.4 mini: 51.4 (#62), MiniMax-M2.5: 38.6 (#132)
| Benchmark | GPT-5.4 mini | MiniMax-M2.5 |
|---|---|---|
| Vectara Hallucination Rate | 5.5% | 9.1% |
| LMArena Expert | 1434 | 1378 |
| GPQA Diamond | 86.9% | — |
| SimpleQA Verified | 29.4% | — |
Multimodal Not comparable
GPT-5.4 mini: 39.0 (#51), MiniMax-M2.5: —
| Benchmark | GPT-5.4 mini | MiniMax-M2.5 |
|---|---|---|
| LMArena Vision | 1245 | — |
Multilingual GPT-5.4 mini leads
GPT-5.4 mini: 52.1 (#94), MiniMax-M2.5: 47.3 (#147)
| Benchmark | GPT-5.4 mini | MiniMax-M2.5 |
|---|---|---|
| LMArena Non-English | 1405 | 1338 |
| LMArena Chinese | 1446 | 1392 |
| LMArena French | 1440 | 1363 |
| LMArena German | 1407 | 1361 |
| LMArena Japanese | 1375 | 1169 |
| LMArena Korean | 1369 | 1232 |
| LMArena Russian | 1417 | 1359 |
| LMArena Spanish | 1403 | 1354 |
Instruction Following GPT-5.4 mini leads
GPT-5.4 mini: 74.3 (#99), MiniMax-M2.5: 71.7 (#143)
| Benchmark | GPT-5.4 mini | MiniMax-M2.5 |
|---|---|---|
| LMArena Instruction Following | 1405 | 1353 |
Long Context GPT-5.4 mini leads
GPT-5.4 mini: 42.9 (#107), MiniMax-M2.5: 37.4 (#211)
| Benchmark | GPT-5.4 mini | MiniMax-M2.5 |
|---|---|---|
| LMArena Longer Query | 1407 | 1367 |
| CL-bench | — | 11.4% |
| CL-bench Life | — | 6.3% |
Writing & Preference GPT-5.4 mini leads
GPT-5.4 mini: 64.2 (#56), MiniMax-M2.5: 54.1 (#149)
| Benchmark | GPT-5.4 mini | MiniMax-M2.5 |
|---|---|---|
| LMArena Text | 1411 | 1359 |
| LMArena Creative Writing | 1370 | 1331 |
| EQ-Bench Creative Writing | 1665 | 1361 |
| LMArena Multi-Turn | 1429 | 1364 |
Frequently asked questions
Is GPT-5.4 mini better than MiniMax-M2.5?
GPT-5.4 mini is the stronger model overall, scoring 44.8 to 38.1 on the Noometry Index. MiniMax-M2.5 costs 3.2× less per token, which makes it the better buy when GPT-5.4 mini's lead doesn't matter for your workload.
Which is cheaper, GPT-5.4 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-5.4 mini lists at $0.75 and $4.50.
Is GPT-5.4 mini or MiniMax-M2.5 better for coding?
MiniMax-M2.5 scores higher on coding benchmarks: 47.9 versus 45.0 in the Noometry coding category.
Which has the bigger context window?
GPT-5.4 mini does, with 400K tokens against 205K.
How many benchmarks do GPT-5.4 mini and MiniMax-M2.5 share?
27 benchmarks have published results for both models. GPT-5.4 mini has 46 scored results on Noometry and MiniMax-M2.5 has 33.