Model comparison
GPT-6 Sol vs MiniMax-M2.7
GPT-6 Sol is the stronger model overall, scoring 61.8 to 37.7 on the Noometry Index. MiniMax-M2.7 costs 7.6× less per token, which makes it the better buy when GPT-6 Sol's lead doesn't matter for your workload.
Last verified . 25 shared benchmarks.
MiniMaxSummary
- They share 25 benchmarks with published results for both. GPT-6 Sol scores higher in 8 categories and MiniMax-M2.7 in 1 category; 6 gaps are clear of the uncertainty.
- The widest gap is in math, where GPT-6 Sol leads 87.2 to 25.9.
- The biggest single-benchmark swing is ProofBench: 83% for GPT-6 Sol and 3% for MiniMax-M2.7.
- MiniMax-M2.7 is cheaper at $0.30 / $1.20 per million input/output tokens, against $2 / $10 for GPT-6 Sol.
- GPT-6 Sol accepts more context: 1.05M tokens versus 205K.
- MiniMax-M2.7 has downloadable open weights; the other is API-only.
Side by side
| GPT-6 Sol | MiniMax-M2.7 | |
|---|---|---|
| Provider | OpenAI | MiniMax |
| Noometry Index | 61.8 | 37.7 |
| Released | 2026-09-22 | 2026-03-18 |
| Weights | Proprietary | Open |
| Context window | 1.05M | 205K |
| Max output | 128K | 131K |
| Input $ / M tokens | $2 | $0.30 |
| Output $ / M tokens | $10 | $1.20 |
| Results tracked | 45 | 30 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-6 Sol leads
GPT-6 Sol: 60.1 (#11), MiniMax-M2.7: 41.8 (#120)
| Benchmark | GPT-6 Sol | MiniMax-M2.7 |
|---|---|---|
| LMArena WebDev | 1688 | 1398 |
| SciCode | 57.6% | 47% |
| LMArena Coding | 1447 | 1454 |
| ALE-Bench | 2,462 | 599.25 |
| DeepSWE | 68.8% | — |
| FrontierCode | 49.3% | — |
| WeirdML | — | 37% |
Agentic & Tool Use GPT-6 Sol leads
GPT-6 Sol: 37.2 (#36), MiniMax-M2.7: 25.1 (#111)
| Benchmark | GPT-6 Sol | MiniMax-M2.7 |
|---|---|---|
| Terminal-Bench | — | 45.1% |
| APEX-Agents | 54.3% | — |
| ExploitBench | — | 13.3% |
| GBAEval | — | 0% |
| GDP.pdf | 26.4% | — |
| Vending-Bench 2 | 14,428 | — |
Reasoning GPT-6 Sol leads
GPT-6 Sol: 74.0 (#9), MiniMax-M2.7: 19.7 (#253)
| Benchmark | GPT-6 Sol | MiniMax-M2.7 |
|---|---|---|
| NYT Connections (extended) | 90.1% | 24.7% |
| CritPt | 30.9% | 0.6% |
| LMArena Hard Prompts | 1418 | 1422 |
| Epoch Capabilities Index | 162.72 | 145.85 |
| ARC-AGI-2 | 89.6% | — |
| ARC-AGI-1 | 95.5% | — |
| Thematic Generalization | — | 39.3% |
| EBR-Bench | 53.3% | — |
| Mystery Game Puzzles | 56% | — |
| DTBench | 97.3% | — |
| LMCA | 59.1% | — |
Math GPT-6 Sol leads
GPT-6 Sol: 87.2 (#7), MiniMax-M2.7: 25.9 (#263)
| Benchmark | GPT-6 Sol | MiniMax-M2.7 |
|---|---|---|
| ProofBench | 83% | 3% |
| LMArena Math | 1402 | 1420 |
| FrontierMath (Tiers 1-3) | 89.8% | — |
| FrontierMath Tier 4 | 90% | — |
| OTIS Mock AIME 2024-2025 | 100% | — |
Knowledge GPT-6 Sol leads
GPT-6 Sol: 64.8 (#15), MiniMax-M2.7: 37.7 (#152)
| Benchmark | GPT-6 Sol | MiniMax-M2.7 |
|---|---|---|
| Vectara Hallucination Rate | 6.5% | 12.9% |
| LMArena Expert | 1439 | 1444 |
| GPQA Diamond | 94.3% | — |
| SimpleQA Verified | 60.7% | — |
Multimodal Not comparable
GPT-6 Sol: 47.6 (#10), MiniMax-M2.7: —
| Benchmark | GPT-6 Sol | MiniMax-M2.7 |
|---|---|---|
| LMArena Vision | 1245 | — |
| Blueprint-Bench 2 | 36.9% | — |
| Furniture Assembly | 58.3% | — |
Multilingual Too close to call
GPT-6 Sol: 50.5 (#118), MiniMax-M2.7: 50.3 (#123)
| Benchmark | GPT-6 Sol | MiniMax-M2.7 |
|---|---|---|
| LMArena Non-English | 1385 | 1382 |
| LMArena Chinese | 1405 | 1441 |
| LMArena French | 1410 | 1421 |
| LMArena German | 1390 | 1398 |
| LMArena Japanese | 1385 | 1262 |
| LMArena Korean | 1341 | 1313 |
| LMArena Russian | 1401 | 1383 |
| LMArena Spanish | 1384 | 1403 |
Instruction Following Too close to call
GPT-6 Sol: 74.5 (#94), MiniMax-M2.7: 74.1 (#103)
| Benchmark | GPT-6 Sol | MiniMax-M2.7 |
|---|---|---|
| LMArena Instruction Following | 1412 | 1405 |
Long Context Too close to call
GPT-6 Sol: 43.1 (#108), MiniMax-M2.7: 43.3 (#99)
| Benchmark | GPT-6 Sol | MiniMax-M2.7 |
|---|---|---|
| LMArena Longer Query | 1411 | 1419 |
Writing & Preference GPT-6 Sol leads
GPT-6 Sol: 71.9 (#18), MiniMax-M2.7: 58.9 (#112)
| Benchmark | GPT-6 Sol | MiniMax-M2.7 |
|---|---|---|
| LMArena Text | 1395 | 1405 |
| LMArena Creative Writing | 1378 | 1354 |
| LMArena Multi-Turn | 1412 | 1412 |
| EQ-Bench Creative Writing | 2125 | — |
Frequently asked questions
Is GPT-6 Sol better than MiniMax-M2.7?
GPT-6 Sol is the stronger model overall, scoring 61.8 to 37.7 on the Noometry Index. MiniMax-M2.7 costs 7.6× less per token, which makes it the better buy when GPT-6 Sol's lead doesn't matter for your workload.
Which is cheaper, GPT-6 Sol or MiniMax-M2.7?
MiniMax-M2.7 is cheaper. It lists at $0.30 per million input tokens and $1.20 per million output tokens; GPT-6 Sol lists at $2 and $10.
Is GPT-6 Sol or MiniMax-M2.7 better for coding?
GPT-6 Sol scores higher on coding benchmarks: 60.1 versus 41.8 in the Noometry coding category.
Which has the bigger context window?
GPT-6 Sol does, with 1.05M tokens against 205K.
How many benchmarks do GPT-6 Sol and MiniMax-M2.7 share?
25 benchmarks have published results for both models. GPT-6 Sol has 45 scored results on Noometry and MiniMax-M2.7 has 30.