Model comparison
Claude 3.5 Haiku vs Kimi K3
Kimi K3 is the stronger model overall, scoring 59.5 to 29.2 on the Noometry Index.
Last verified . 25 shared benchmarks.
Summary
- They share 25 benchmarks with published results for both. Claude 3.5 Haiku scores higher in 0 categories and Kimi K3 in 10 categories; 10 gaps are clear of the uncertainty.
- The widest gap is in math, where Kimi K3 leads 74.2 to 14.7.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 4.3% for Claude 3.5 Haiku and 97.2% for Kimi K3.
- Kimi K3 has downloadable open weights; the other is API-only.
Side by side
| Claude 3.5 Haiku | Kimi K3 | |
|---|---|---|
| Provider | Anthropic | Moonshot AI |
| Noometry Index | 29.2 | 59.5 |
| Released | 2024-10-22 | 2026-07-16 |
| Weights | Proprietary | Open |
| Context window | — | 1.05M |
| Max output | — | 1.05M |
| Input $ / M tokens | — | $3 |
| Output $ / M tokens | — | $15 |
| Results tracked | 49 | 53 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Kimi K3 leads
Claude 3.5 Haiku: 32.9 (#265), Kimi K3: 61.0 (#10)
| Benchmark | Claude 3.5 Haiku | Kimi K3 |
|---|---|---|
| SciCode | 27.4% | 59.5% |
| WeirdML | 30.7% | 82.6% |
| LMArena Coding | 1286 | 1508 |
| DeepSWE | — | 68.5% |
| FrontierCode | — | 44.2% |
| Aider Polyglot | 28% | — |
| LMArena WebDev | — | 1654 |
| FrontierSWE | — | 25.9% |
| BigCodeBench Instruct | 46.1% | — |
| LiveBench Coding | 51.4% | — |
| BigCodeBench Complete | 59% | — |
| CadEval | 32% | — |
| ALE-Bench | — | 1,524 |
Agentic & Tool Use Kimi K3 leads
Claude 3.5 Haiku: 28.0 (#95), Kimi K3: 41.8 (#20)
| Benchmark | Claude 3.5 Haiku | Kimi K3 |
|---|---|---|
| APEX-Agents | — | 50.6% |
| τ²-bench Banking | — | 37.1% |
| PostTrainBench | — | 32% |
| BALROG | 19.3% | — |
| GBAEval | — | 48.3% |
| GDP.pdf | — | 19% |
| Vending-Bench 2 | — | 5,165 |
Reasoning Kimi K3 leads
Claude 3.5 Haiku: 17.7 (#290), Kimi K3: 63.0 (#17)
| Benchmark | Claude 3.5 Haiku | Kimi K3 |
|---|---|---|
| CritPt | 0% | 23.4% |
| LMArena Hard Prompts | 1251 | 1496 |
| DTBench | 56.7% | 91.2% |
| Epoch Capabilities Index | 127.15 | 157.45 |
| ARC-AGI-2 | — | 60.4% |
| SimpleBench | — | 60.7% |
| NYT Connections (extended) | — | 93.6% |
| ARC-AGI-1 | — | 94.5% |
| Chess Puzzles | — | 39% |
| LiveBench Reasoning | 28.1% | — |
| Mystery Game Puzzles | — | 26% |
| LiveBench Data Analysis | 48.5% | — |
| LMCA | — | 52.7% |
| Surface Evolver Bench | — | 95% |
| ForecastBench | — | 61.1 |
| LiveBench | 43.5% | — |
Math Kimi K3 leads
Claude 3.5 Haiku: 14.7 (#300), Kimi K3: 74.2 (#16)
| Benchmark | Claude 3.5 Haiku | Kimi K3 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 4.3% | 97.2% |
| LMArena Math | 1244 | 1491 |
| FrontierMath (Tiers 1-3) | — | 72.2% |
| FrontierMath Tier 4 | — | 39% |
| MathArena Final-Answer Competitions | — | 87.8% |
| ProofBench | — | 87% |
| Omni-MATH | 22.4% | — |
| LiveBench Math | 35.5% | — |
| MATH Level 5 | 46.4% | — |
| FrontierMath (Feb 2025 set) | 0.3% | — |
Knowledge Kimi K3 leads
Claude 3.5 Haiku: 18.7 (#281), Kimi K3: 63.2 (#21)
| Benchmark | Claude 3.5 Haiku | Kimi K3 |
|---|---|---|
| GPQA Diamond | 38.1% | 93.1% |
| LMArena Expert | 1208 | 1521 |
| SimpleQA Verified | — | 50.6% |
| MMLU-Pro | 60.5% | — |
| Confabulations | 36.7% | — |
| GPQA (HELM) | 36.3% | — |
| MMLU | 74.3% | — |
Multimodal Kimi K3 leads
Claude 3.5 Haiku: 26.8 (#117), Kimi K3: 37.8 (#70)
| Benchmark | Claude 3.5 Haiku | Kimi K3 |
|---|---|---|
| LMArena Vision | 1092 | — |
| GeoBench | 34% | — |
| Blueprint-Bench 2 | — | 29.5% |
| Furniture Assembly | — | 34.2% |
Multilingual Kimi K3 leads
Claude 3.5 Haiku: 40.0 (#218), Kimi K3: 56.3 (#21)
| Benchmark | Claude 3.5 Haiku | Kimi K3 |
|---|---|---|
| LMArena Non-English | 1238 | 1466 |
| LMArena Chinese | 1229 | 1529 |
| LMArena French | 1264 | 1491 |
| LMArena German | 1237 | 1488 |
| LMArena Japanese | 1175 | 1487 |
| LMArena Korean | 1173 | 1458 |
| LMArena Russian | 1253 | 1482 |
| LMArena Spanish | 1261 | 1472 |
Instruction Following Kimi K3 leads
Claude 3.5 Haiku: 62.9 (#234), Kimi K3: 77.7 (#14)
| Benchmark | Claude 3.5 Haiku | Kimi K3 |
|---|---|---|
| LMArena Instruction Following | 1241 | 1483 |
| LiveBench Instruction Following | 61.9% | — |
| IFEval | 79.2% | — |
Long Context Kimi K3 leads
Claude 3.5 Haiku: 38.3 (#200), Kimi K3: 45.8 (#29)
| Benchmark | Claude 3.5 Haiku | Kimi K3 |
|---|---|---|
| LMArena Longer Query | 1261 | 1494 |
Writing & Preference Kimi K3 leads
Claude 3.5 Haiku: 42.7 (#234), Kimi K3: 76.6 (#4)
| Benchmark | Claude 3.5 Haiku | Kimi K3 |
|---|---|---|
| LMArena Text | 1255 | 1476 |
| LMArena Creative Writing | 1233 | 1454 |
| EQ-Bench Creative Writing | 1146 | 2082 |
| LMArena Multi-Turn | 1265 | 1488 |
| Short-Story Creative Writing | 73.5% | — |
| WildBench | 76% | — |
| EQ-Bench 4 | — | 1339 |
| LiveBench Language | 35.4% | — |
Frequently asked questions
Is Claude 3.5 Haiku better than Kimi K3?
Kimi K3 is the stronger model overall, scoring 59.5 to 29.2 on the Noometry Index.
Is Claude 3.5 Haiku or Kimi K3 better for coding?
Kimi K3 scores higher on coding benchmarks: 61.0 versus 32.9 in the Noometry coding category.
How many benchmarks do Claude 3.5 Haiku and Kimi K3 share?
25 benchmarks have published results for both models. Claude 3.5 Haiku has 49 scored results on Noometry and Kimi K3 has 53.