Model comparison
Kimi K2 (Jul 2025) vs MiniMax-M3
MiniMax-M3 is the stronger model overall, scoring 43.5 to 40.9 on the Noometry Index.
Last verified . 21 shared benchmarks.
MiniMaxSummary
- They share 21 benchmarks with published results for both. Kimi K2 (Jul 2025) scores higher in 4 categories and MiniMax-M3 in 5 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where MiniMax-M3 leads 58.4 to 36.9.
- The biggest single-benchmark swing is SimpleBench: 26.3% for Kimi K2 (Jul 2025) and 45.8% for MiniMax-M3.
- MiniMax-M3 is cheaper at $0.30 / $1.20 per million input/output tokens, against $0.57 / $2.30 for Kimi K2 (Jul 2025).
- MiniMax-M3 accepts more context: 1M tokens versus 262K.
Side by side
| Kimi K2 (Jul 2025) | MiniMax-M3 | |
|---|---|---|
| Provider | Moonshot AI | MiniMax |
| Noometry Index | 40.9 | 43.5 |
| Released | 2025-07-12 | 2026-06-01 |
| Weights | Open | Open |
| Context window | 262K | 1M |
| Max output | 262K | 512K |
| Input $ / M tokens | $0.57 | $0.30 |
| Output $ / M tokens | $2.30 | $1.20 |
| Results tracked | 42 | 41 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Kimi K2 (Jul 2025) leads
Kimi K2 (Jul 2025): 43.1 (#83), MiniMax-M3: 41.5 (#111)
| Benchmark | Kimi K2 (Jul 2025) | MiniMax-M3 |
|---|---|---|
| LMArena Coding | 1399 | 1470 |
| ALE-Bench | 597.5 | 640.02 |
| FrontierCode | — | 14.7% |
| SWE-bench Verified (bash only) | 63.4% | — |
| Aider Polyglot | 59.1% | — |
| LMArena WebDev | — | 1482 |
| SciCode | — | 47.1% |
| GSO | 4.9% | — |
| WeirdML | 42.8% | — |
Agentic & Tool Use Kimi K2 (Jul 2025) leads
Kimi K2 (Jul 2025): 32.2 (#59), MiniMax-M3: 22.5 (#121)
| Benchmark | Kimi K2 (Jul 2025) | MiniMax-M3 |
|---|---|---|
| Terminal-Bench | 35.7% | — |
| APEX-Agents | — | 37.7% |
| Berkeley Function Calling Leaderboard | 59.1% | — |
| OSWorld 2.0 | — | 4.6% |
| GBAEval | — | 0.9% |
| METR Time Horizons | 59.2% | — |
| Vending-Bench 2 | — | 2,158 |
Reasoning MiniMax-M3 leads
Kimi K2 (Jul 2025): 22.7 (#173), MiniMax-M3: 29.3 (#83)
| Benchmark | Kimi K2 (Jul 2025) | MiniMax-M3 |
|---|---|---|
| SimpleBench | 26.3% | 45.8% |
| LMArena Hard Prompts | 1384 | 1447 |
| Epoch Capabilities Index | 146.02 | 147 |
| ForecastBench | 60.2 | 61.4 |
| Kagi LLM Benchmark | 64.4% | — |
| NYT Connections (extended) | — | 65.1% |
| CritPt | — | 3.7% |
| Chess Puzzles | — | 14% |
| Mystery Game Puzzles | — | 8% |
| DTBench | — | 78.9% |
| LMCA | — | 33.7% |
| Surface Evolver Bench | — | 55% |
Math Too close to call
Kimi K2 (Jul 2025): 40.1 (#81), MiniMax-M3: 39.1 (#83)
| Benchmark | Kimi K2 (Jul 2025) | MiniMax-M3 |
|---|---|---|
| LMArena Math | 1396 | 1432 |
| OTIS Mock AIME 2024-2025 | — | 71.1% |
| ProofBench | — | 18% |
| Omni-MATH | 65.4% | — |
| FrontierMath (Feb 2025 set) | 21.4% | — |
| FrontierMath Tier 4 (v1) | 0% | — |
Knowledge MiniMax-M3 leads
Kimi K2 (Jul 2025): 36.9 (#153), MiniMax-M3: 58.4 (#32)
| Benchmark | Kimi K2 (Jul 2025) | MiniMax-M3 |
|---|---|---|
| LMArena Expert | 1363 | 1462 |
| GPQA Diamond | — | 90.9% |
| MMLU-Pro | 81.9% | — |
| Confabulations | 20.4% | — |
| Vectara Hallucination Rate | 17.9% | — |
| GPQA (HELM) | 65.3% | — |
Multimodal Not comparable
Kimi K2 (Jul 2025): —, MiniMax-M3: 39.5 (#45)
| Benchmark | Kimi K2 (Jul 2025) | MiniMax-M3 |
|---|---|---|
| LMArena Vision | — | 1254 |
| LMArena Document | — | 1435 |
Multilingual MiniMax-M3 leads
Kimi K2 (Jul 2025): 49.8 (#125), MiniMax-M3: 53.2 (#70)
| Benchmark | Kimi K2 (Jul 2025) | MiniMax-M3 |
|---|---|---|
| LMArena Non-English | 1372 | 1420 |
| LMArena Chinese | 1416 | 1465 |
| LMArena French | 1379 | 1445 |
| LMArena German | 1386 | 1424 |
| LMArena Japanese | 1349 | 1381 |
| LMArena Korean | 1324 | 1372 |
| LMArena Russian | 1384 | 1430 |
| LMArena Spanish | 1383 | 1431 |
Instruction Following MiniMax-M3 leads
Kimi K2 (Jul 2025): 71.3 (#151), MiniMax-M3: 75.7 (#56)
| Benchmark | Kimi K2 (Jul 2025) | MiniMax-M3 |
|---|---|---|
| LMArena Instruction Following | 1347 | 1433 |
| IFEval | 85% | — |
Long Context MiniMax-M3 leads
Kimi K2 (Jul 2025): 41.2 (#140), MiniMax-M3: 44.2 (#68)
| Benchmark | Kimi K2 (Jul 2025) | MiniMax-M3 |
|---|---|---|
| LMArena Longer Query | 1353 | 1445 |
| Fiction.LiveBench | 66.7% | — |
| CL-bench | 17.6% | — |
Writing & Preference Too close to call
Kimi K2 (Jul 2025): 62.5 (#76), MiniMax-M3: 62.3 (#79)
| Benchmark | Kimi K2 (Jul 2025) | MiniMax-M3 |
|---|---|---|
| LMArena Text | 1380 | 1433 |
| LMArena Creative Writing | 1349 | 1405 |
| LMArena Multi-Turn | 1371 | 1441 |
| Short-Story Creative Writing | 85.6% | — |
| EQ-Bench Creative Writing | 1666 | — |
| WildBench | 86.2% | — |
| EQ-Bench 4 | — | 1150 |
Frequently asked questions
Is Kimi K2 (Jul 2025) better than MiniMax-M3?
MiniMax-M3 is the stronger model overall, scoring 43.5 to 40.9 on the Noometry Index.
Which is cheaper, Kimi K2 (Jul 2025) or MiniMax-M3?
MiniMax-M3 is cheaper. It lists at $0.30 per million input tokens and $1.20 per million output tokens; Kimi K2 (Jul 2025) lists at $0.57 and $2.30.
Is Kimi K2 (Jul 2025) or MiniMax-M3 better for coding?
Kimi K2 (Jul 2025) scores higher on coding benchmarks: 43.1 versus 41.5 in the Noometry coding category.
Which has the bigger context window?
MiniMax-M3 does, with 1M tokens against 262K.
How many benchmarks do Kimi K2 (Jul 2025) and MiniMax-M3 share?
21 benchmarks have published results for both models. Kimi K2 (Jul 2025) has 42 scored results on Noometry and MiniMax-M3 has 41.