Model comparison
DeepSeek LLM 67B vs MiniMax-M2
MiniMax-M2 is the stronger model overall, scoring 37.4 to 24.9 on the Noometry Index.
Last verified . 10 shared benchmarks.
DeepSeek
MiniMaxSummary
- They share 10 benchmarks with published results for both. DeepSeek LLM 67B scores higher in 0 categories and MiniMax-M2 in 8 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where MiniMax-M2 leads 37.0 to 7.0.
Side by side
| DeepSeek LLM 67B | MiniMax-M2 | |
|---|---|---|
| Provider | DeepSeek | MiniMax |
| Noometry Index | 24.9 | 37.4 |
| Released | 2023-11-29 | 2025-10-27 |
| Weights | Open | Open |
| Context window | — | 205K |
| Max output | — | 131K |
| Input $ / M tokens | — | $0.30 |
| Output $ / M tokens | — | $1.20 |
| Results tracked | 15 | 21 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding MiniMax-M2 leads
DeepSeek LLM 67B: 31.9 (#278), MiniMax-M2: 39.3 (#159)
| Benchmark | DeepSeek LLM 67B | MiniMax-M2 |
|---|---|---|
| LMArena Coding | 1096 | 1370 |
| SWE-bench Verified (bash only) | — | 61% |
| LMArena WebDev | — | 1297 |
Agentic & Tool Use Not comparable
DeepSeek LLM 67B: —, MiniMax-M2: 25.1 (#109)
| Benchmark | DeepSeek LLM 67B | MiniMax-M2 |
|---|---|---|
| Terminal-Bench | — | 30% |
| Vending-Bench 2 | — | 160.6 |
Reasoning MiniMax-M2 leads
DeepSeek LLM 67B: 16.5 (#304), MiniMax-M2: 19.4 (#258)
| Benchmark | DeepSeek LLM 67B | MiniMax-M2 |
|---|---|---|
| LMArena Hard Prompts | 1070 | 1357 |
| Kagi LLM Benchmark | — | 57.8% |
| NYT Connections (extended) | — | 14.8% |
| Chess Puzzles | 0% | — |
| Epoch Capabilities Index | 110.5 | — |
Math MiniMax-M2 leads
DeepSeek LLM 67B: 8.7 (#324), MiniMax-M2: 37.3 (#160)
| Benchmark | DeepSeek LLM 67B | MiniMax-M2 |
|---|---|---|
| LMArena Math | 1108 | 1352 |
| OTIS Mock AIME 2024-2025 | 0.8% | — |
| MATH Level 5 | 6.4% | — |
Knowledge MiniMax-M2 leads
DeepSeek LLM 67B: 7.0 (#313), MiniMax-M2: 37.0 (#163)
| Benchmark | DeepSeek LLM 67B | MiniMax-M2 |
|---|---|---|
| GPQA Diamond | 24.6% | — |
| LMArena Expert | — | 1337 |
Multilingual MiniMax-M2 leads
DeepSeek LLM 67B: 29.4 (#267), MiniMax-M2: 45.3 (#171)
| Benchmark | DeepSeek LLM 67B | MiniMax-M2 |
|---|---|---|
| LMArena Non-English | 1073 | 1313 |
| LMArena Chinese | 1132 | 1366 |
| LMArena French | — | 1335 |
| LMArena German | — | 1355 |
| LMArena Russian | — | 1331 |
| LMArena Spanish | — | 1326 |
Instruction Following MiniMax-M2 leads
DeepSeek LLM 67B: 55.4 (#277), MiniMax-M2: 70.2 (#166)
| Benchmark | DeepSeek LLM 67B | MiniMax-M2 |
|---|---|---|
| LMArena Instruction Following | 1079 | 1328 |
Long Context MiniMax-M2 leads
DeepSeek LLM 67B: 33.1 (#265), MiniMax-M2: 40.5 (#153)
| Benchmark | DeepSeek LLM 67B | MiniMax-M2 |
|---|---|---|
| LMArena Longer Query | 1092 | 1331 |
Writing & Preference MiniMax-M2 leads
DeepSeek LLM 67B: 31.6 (#282), MiniMax-M2: 53.0 (#162)
| Benchmark | DeepSeek LLM 67B | MiniMax-M2 |
|---|---|---|
| LMArena Text | 1105 | 1340 |
| LMArena Creative Writing | 1067 | 1286 |
| LMArena Multi-Turn | 1082 | 1361 |
Frequently asked questions
Is DeepSeek LLM 67B better than MiniMax-M2?
MiniMax-M2 is the stronger model overall, scoring 37.4 to 24.9 on the Noometry Index.
Is DeepSeek LLM 67B or MiniMax-M2 better for coding?
MiniMax-M2 scores higher on coding benchmarks: 39.3 versus 31.9 in the Noometry coding category.
How many benchmarks do DeepSeek LLM 67B and MiniMax-M2 share?
10 benchmarks have published results for both models. DeepSeek LLM 67B has 15 scored results on Noometry and MiniMax-M2 has 21.