Model comparison
DeepSeek-V3.2-Speciale vs MiniMax-M2.7
DeepSeek-V3.2-Speciale is the stronger model overall, scoring 39.1 to 37.4 on the Noometry Index. MiniMax-M2.7 costs 1.6× less per token, which makes it the better buy when DeepSeek-V3.2-Speciale's lead doesn't matter for your workload.
Last verified . 1 shared benchmarks.
DeepSeek
MiniMaxSummary
- They share 1 benchmark with published results for both. DeepSeek-V3.2-Speciale scores higher in 1 category and MiniMax-M2.7 in 2 categories; 3 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where DeepSeek-V3.2-Speciale leads 32.1 to 19.0.
- The biggest single-benchmark swing is WeirdML: 46.7% for DeepSeek-V3.2-Speciale and 37% for MiniMax-M2.7.
- MiniMax-M2.7 is cheaper at $0.30 / $1.20 per million input/output tokens, against $0.58 / $1.68 for DeepSeek-V3.2-Speciale.
- MiniMax-M2.7 accepts more context: 205K tokens versus 128K.
Side by side
| DeepSeek-V3.2-Speciale | MiniMax-M2.7 | |
|---|---|---|
| Provider | DeepSeek | MiniMax |
| Noometry Index | 39.1 | 37.4 |
| Released | 2025-12-01 | 2026-03-18 |
| Weights | Open | Open |
| Context window | 128K | 205K |
| Max output | 128K | 131K |
| Input $ / M tokens | $0.58 | $0.30 |
| Output $ / M tokens | $1.68 | $1.20 |
| Results tracked | 3 | 30 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding MiniMax-M2.7 leads
DeepSeek-V3.2-Speciale: 39.9 (#136), MiniMax-M2.7: 41.4 (#115)
| Benchmark | DeepSeek-V3.2-Speciale | MiniMax-M2.7 |
|---|---|---|
| WeirdML | 46.7% | 37% |
| LMArena WebDev | — | 1397 |
| SciCode | — | 47% |
| LMArena Coding | — | 1454 |
| ALE-Bench | — | 599.25 |
Agentic & Tool Use Not comparable
DeepSeek-V3.2-Speciale: —, MiniMax-M2.7: 24.9 (#103)
| Benchmark | DeepSeek-V3.2-Speciale | MiniMax-M2.7 |
|---|---|---|
| Terminal-Bench | — | 45.1% |
| ExploitBench | — | 13.3% |
| GBAEval | — | 0% |
Reasoning DeepSeek-V3.2-Speciale leads
DeepSeek-V3.2-Speciale: 32.1 (#65), MiniMax-M2.7: 19.0 (#248)
| Benchmark | DeepSeek-V3.2-Speciale | MiniMax-M2.7 |
|---|---|---|
| SimpleBench | 52.6% | — |
| NYT Connections (extended) | — | 24.7% |
| CritPt | — | 0.6% |
| Thematic Generalization | — | 39.3% |
| LMArena Hard Prompts | — | 1422 |
| Epoch Capabilities Index | — | 145.84 |
Math Not comparable
DeepSeek-V3.2-Speciale: —, MiniMax-M2.7: 24.9 (#254)
| Benchmark | DeepSeek-V3.2-Speciale | MiniMax-M2.7 |
|---|---|---|
| ProofBench | — | 3% |
| LMArena Math | — | 1420 |
Knowledge Not comparable
DeepSeek-V3.2-Speciale: —, MiniMax-M2.7: 37.1 (#151)
| Benchmark | DeepSeek-V3.2-Speciale | MiniMax-M2.7 |
|---|---|---|
| Vectara Hallucination Rate | — | 12.9% |
| LMArena Expert | — | 1443 |
Multilingual Not comparable
DeepSeek-V3.2-Speciale: —, MiniMax-M2.7: 50.5 (#120)
| Benchmark | DeepSeek-V3.2-Speciale | MiniMax-M2.7 |
|---|---|---|
| LMArena Non-English | — | 1382 |
| LMArena Chinese | — | 1439 |
| LMArena French | — | 1422 |
| LMArena German | — | 1398 |
| LMArena Japanese | — | 1261 |
| LMArena Korean | — | 1313 |
| LMArena Russian | — | 1383 |
| LMArena Spanish | — | 1403 |
Instruction Following Not comparable
DeepSeek-V3.2-Speciale: —, MiniMax-M2.7: 74.2 (#101)
| Benchmark | DeepSeek-V3.2-Speciale | MiniMax-M2.7 |
|---|---|---|
| LMArena Instruction Following | — | 1404 |
Long Context Not comparable
DeepSeek-V3.2-Speciale: —, MiniMax-M2.7: 43.3 (#94)
| Benchmark | DeepSeek-V3.2-Speciale | MiniMax-M2.7 |
|---|---|---|
| LMArena Longer Query | — | 1419 |
Writing & Preference MiniMax-M2.7 leads
DeepSeek-V3.2-Speciale: 46.1 (#217), MiniMax-M2.7: 59.1 (#107)
| Benchmark | DeepSeek-V3.2-Speciale | MiniMax-M2.7 |
|---|---|---|
| LMArena Text | — | 1404 |
| LMArena Creative Writing | — | 1354 |
| EQ-Bench Creative Writing | 1276 | — |
| LMArena Multi-Turn | — | 1413 |
Frequently asked questions
Is DeepSeek-V3.2-Speciale better than MiniMax-M2.7?
DeepSeek-V3.2-Speciale is the stronger model overall, scoring 39.1 to 37.4 on the Noometry Index. MiniMax-M2.7 costs 1.6× less per token, which makes it the better buy when DeepSeek-V3.2-Speciale's lead doesn't matter for your workload.
Which is cheaper, DeepSeek-V3.2-Speciale 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; DeepSeek-V3.2-Speciale lists at $0.58 and $1.68.
Is DeepSeek-V3.2-Speciale or MiniMax-M2.7 better for coding?
MiniMax-M2.7 scores higher on coding benchmarks: 41.4 versus 39.9 in the Noometry coding category.
Which has the bigger context window?
MiniMax-M2.7 does, with 205K tokens against 128K.
How many benchmarks do DeepSeek-V3.2-Speciale and MiniMax-M2.7 share?
1 benchmark has published results for both models. DeepSeek-V3.2-Speciale has 3 scored results on Noometry and MiniMax-M2.7 has 30.