Model comparison
Kimi K3 vs MiMo-V2.6-Pro
Kimi K3 is the stronger model overall, scoring 59.5 to 50.3 on the Noometry Index. MiMo-V2.6-Pro costs 11× less per token, which makes it the better buy when Kimi K3's lead doesn't matter for your workload.
Last verified . 18 shared benchmarks.
Summary
- They share 18 benchmarks with published results for both. Kimi K3 scores higher in 6 categories and MiMo-V2.6-Pro in 4 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Kimi K3 leads 63.0 to 43.1.
- The biggest single-benchmark swing is ProofBench: 87% for Kimi K3 and 70% for MiMo-V2.6-Pro.
- MiMo-V2.6-Pro is cheaper at $0.43 / $0.87 per million input/output tokens, against $3 / $15 for Kimi K3.
Side by side
| Kimi K3 | MiMo-V2.6-Pro | |
|---|---|---|
| Provider | Moonshot AI | Xiaomi |
| Noometry Index | 59.5 | 50.3 |
| Released | 2026-07-16 | 2026-09-21 |
| Weights | Open | Open |
| Context window | 1.05M | 1.05M |
| Max output | 1.05M | 131K |
| Input $ / M tokens | $3 | $0.43 |
| Output $ / M tokens | $15 | $0.87 |
| Results tracked | 53 | 19 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Kimi K3 leads
Kimi K3: 61.0 (#10), MiMo-V2.6-Pro: 55.5 (#23)
| Benchmark | Kimi K3 | MiMo-V2.6-Pro |
|---|---|---|
| LMArena WebDev | 1654 | 1629 |
| SciCode | 59.5% | 60.9% |
| LMArena Coding | 1508 | 1534 |
| ALE-Bench | 1,524 | 1,158 |
| DeepSWE | 68.5% | — |
| FrontierCode | 44.2% | — |
| FrontierSWE | 25.9% | — |
| WeirdML | 82.6% | — |
Agentic & Tool Use Kimi K3 leads
Kimi K3: 41.8 (#20), MiMo-V2.6-Pro: 37.5 (#35)
| Benchmark | Kimi K3 | MiMo-V2.6-Pro |
|---|---|---|
| APEX-Agents | 50.6% | 59.5% |
| τ²-bench Banking | 37.1% | — |
| PostTrainBench | 32% | — |
| GBAEval | 48.3% | — |
| GDP.pdf | 19% | — |
| Vending-Bench 2 | 5,165 | — |
Reasoning Kimi K3 leads
Kimi K3: 63.0 (#17), MiMo-V2.6-Pro: 43.1 (#50)
| Benchmark | Kimi K3 | MiMo-V2.6-Pro |
|---|---|---|
| CritPt | 23.4% | 26.6% |
| LMArena Hard Prompts | 1496 | 1512 |
| ARC-AGI-2 | 60.4% | — |
| SimpleBench | 60.7% | — |
| NYT Connections (extended) | 93.6% | — |
| ARC-AGI-1 | 94.5% | — |
| Chess Puzzles | 39% | — |
| Mystery Game Puzzles | 26% | — |
| DTBench | 91.2% | — |
| LMCA | 52.7% | — |
| Surface Evolver Bench | 95% | — |
| Epoch Capabilities Index | 157.45 | — |
| ForecastBench | 61.1 | — |
Math Kimi K3 leads
Kimi K3: 74.2 (#16), MiMo-V2.6-Pro: 54.5 (#45)
| Benchmark | Kimi K3 | MiMo-V2.6-Pro |
|---|---|---|
| ProofBench | 87% | 70% |
| LMArena Math | 1491 | 1494 |
| FrontierMath (Tiers 1-3) | 72.2% | — |
| FrontierMath Tier 4 | 39% | — |
| MathArena Final-Answer Competitions | 87.8% | — |
| OTIS Mock AIME 2024-2025 | 97.2% | — |
Knowledge Kimi K3 leads
Kimi K3: 63.2 (#21), MiMo-V2.6-Pro: 43.5 (#92)
| Benchmark | Kimi K3 | MiMo-V2.6-Pro |
|---|---|---|
| LMArena Expert | 1521 | 1543 |
| GPQA Diamond | 93.1% | — |
| SimpleQA Verified | 50.6% | — |
Multimodal MiMo-V2.6-Pro leads
Kimi K3: 37.8 (#70), MiMo-V2.6-Pro: 40.8 (#43)
| Benchmark | Kimi K3 | MiMo-V2.6-Pro |
|---|---|---|
| LMArena Vision | — | 1264 |
| Blueprint-Bench 2 | 29.5% | — |
| Furniture Assembly | 34.2% | — |
Multilingual Too close to call
Kimi K3: 56.3 (#21), MiMo-V2.6-Pro: 56.9 (#14)
| Benchmark | Kimi K3 | MiMo-V2.6-Pro |
|---|---|---|
| LMArena Non-English | 1466 | 1474 |
| LMArena Chinese | 1529 | 1529 |
| LMArena Russian | 1482 | 1480 |
| LMArena French | 1491 | — |
| LMArena German | 1488 | — |
| LMArena Japanese | 1487 | — |
| LMArena Korean | 1458 | — |
| LMArena Spanish | 1472 | — |
Instruction Following Too close to call
Kimi K3: 77.7 (#14), MiMo-V2.6-Pro: 78.2 (#12)
| Benchmark | Kimi K3 | MiMo-V2.6-Pro |
|---|---|---|
| LMArena Instruction Following | 1483 | 1493 |
Long Context Too close to call
Kimi K3: 45.8 (#29), MiMo-V2.6-Pro: 46.0 (#27)
| Benchmark | Kimi K3 | MiMo-V2.6-Pro |
|---|---|---|
| LMArena Longer Query | 1494 | 1501 |
Writing & Preference Kimi K3 leads
Kimi K3: 76.6 (#4), MiMo-V2.6-Pro: 66.8 (#33)
| Benchmark | Kimi K3 | MiMo-V2.6-Pro |
|---|---|---|
| LMArena Text | 1476 | 1492 |
| LMArena Creative Writing | 1454 | 1468 |
| LMArena Multi-Turn | 1488 | 1464 |
| EQ-Bench Creative Writing | 2082 | — |
| EQ-Bench 4 | 1339 | — |
Frequently asked questions
Is Kimi K3 better than MiMo-V2.6-Pro?
Kimi K3 is the stronger model overall, scoring 59.5 to 50.3 on the Noometry Index. MiMo-V2.6-Pro costs 11× less per token, which makes it the better buy when Kimi K3's lead doesn't matter for your workload.
Which is cheaper, Kimi K3 or MiMo-V2.6-Pro?
MiMo-V2.6-Pro is cheaper. It lists at $0.43 per million input tokens and $0.87 per million output tokens; Kimi K3 lists at $3 and $15.
Is Kimi K3 or MiMo-V2.6-Pro better for coding?
Kimi K3 scores higher on coding benchmarks: 61.0 versus 55.5 in the Noometry coding category.
Which has the bigger context window?
Both accept 1.05M tokens.
How many benchmarks do Kimi K3 and MiMo-V2.6-Pro share?
18 benchmarks have published results for both models. Kimi K3 has 53 scored results on Noometry and MiMo-V2.6-Pro has 19.