Model comparison
Granite 4.0 H Small vs Kimi K2.5
Kimi K2.5 is the stronger model overall, scoring 47.8 to 35.9 on the Noometry Index.
Last verified . 14 shared benchmarks.
Summary
- They share 14 benchmarks with published results for both. Granite 4.0 H Small scores higher in 0 categories and Kimi K2.5 in 8 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where Kimi K2.5 leads 53.6 to 31.5.
- The biggest single-benchmark swing is Vectara Hallucination Rate: 5.2% for Granite 4.0 H Small and 14.2% for Kimi K2.5.
Side by side
| Granite 4.0 H Small | Kimi K2.5 | |
|---|---|---|
| Provider | IBM | Moonshot AI |
| Noometry Index | 35.9 | 47.8 |
| Released | — | 2026-01-27 |
| Weights | Open | Open |
| Context window | — | 262K |
| Max output | — | 262K |
| Input $ / M tokens | — | $0.45 |
| Output $ / M tokens | — | $2.25 |
| Results tracked | 19 | 51 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Kimi K2.5 leads
Granite 4.0 H Small: 35.8 (#207), Kimi K2.5: 48.7 (#48)
| Benchmark | Granite 4.0 H Small | Kimi K2.5 |
|---|---|---|
| LMArena Coding | 1247 | 1474 |
| SWE-bench Verified | — | 73.8% |
| SWE-bench Verified (bash only) | — | 70.8% |
| LMArena WebDev | — | 1436 |
| SWE-bench Multilingual | — | 67.3% |
| SciCode | — | 49% |
| WeirdML | — | 45.6% |
| ALE-Bench | — | 821.65 |
Agentic & Tool Use Not comparable
Granite 4.0 H Small: —, Kimi K2.5: 34.0 (#44)
| Benchmark | Granite 4.0 H Small | Kimi K2.5 |
|---|---|---|
| Terminal-Bench | — | 43.2% |
| OSWorld | — | 63.3% |
| Vending-Bench 2 | — | 1,198 |
Reasoning Kimi K2.5 leads
Granite 4.0 H Small: 23.4 (#164), Kimi K2.5: 31.2 (#74)
| Benchmark | Granite 4.0 H Small | Kimi K2.5 |
|---|---|---|
| LMArena Hard Prompts | 1240 | 1454 |
| ARC-AGI-2 | — | 11.8% |
| SimpleBench | — | 46.8% |
| Kagi LLM Benchmark | — | 78.5% |
| NYT Connections (extended) | — | 69.9% |
| ARC-AGI-1 | — | 65.3% |
| CritPt | — | 3.1% |
| Chess Puzzles | — | 12% |
| EnigmaEval | — | 3.4% |
| Thematic Generalization | — | 69.4% |
| Epoch Capabilities Index | — | 148.05 |
Math Kimi K2.5 leads
Granite 4.0 H Small: 29.3 (#219), Kimi K2.5: 50.3 (#46)
| Benchmark | Granite 4.0 H Small | Kimi K2.5 |
|---|---|---|
| LMArena Math | 1251 | 1470 |
| MathArena Final-Answer Competitions | — | 62.3% |
| OTIS Mock AIME 2024-2025 | — | 92.2% |
| Omni-MATH | 29.6% | — |
| FrontierMath (Feb 2025 set) | — | 27.9% |
| FrontierMath Tier 4 (v1) | — | 4.2% |
Knowledge Kimi K2.5 leads
Granite 4.0 H Small: 31.5 (#210), Kimi K2.5: 53.6 (#52)
| Benchmark | Granite 4.0 H Small | Kimi K2.5 |
|---|---|---|
| Vectara Hallucination Rate | 5.2% | 14.2% |
| LMArena Expert | 1244 | 1466 |
| GPQA Diamond | — | 87.6% |
| Humanity's Last Exam | — | 24.4% |
| SimpleQA Verified | — | 34.3% |
| MMLU-Pro | 56.9% | — |
| GPQA (HELM) | 38.3% | — |
Multimodal Not comparable
Granite 4.0 H Small: —, Kimi K2.5: 40.4 (#37)
| Benchmark | Granite 4.0 H Small | Kimi K2.5 |
|---|---|---|
| LMArena Vision | — | 1269 |
| LMArena Document | — | 1430 |
Multilingual Kimi K2.5 leads
Granite 4.0 H Small: 38.7 (#223), Kimi K2.5: 54.1 (#50)
| Benchmark | Granite 4.0 H Small | Kimi K2.5 |
|---|---|---|
| LMArena Non-English | 1216 | 1433 |
| LMArena Chinese | 1250 | 1495 |
| LMArena Russian | 1198 | 1436 |
| LMArena Spanish | 1258 | 1450 |
| LMArena French | — | 1454 |
| LMArena German | — | 1441 |
| LMArena Japanese | — | 1422 |
| LMArena Korean | — | 1412 |
Instruction Following Kimi K2.5 leads
Granite 4.0 H Small: 68.9 (#178), Kimi K2.5: 75.5 (#60)
| Benchmark | Granite 4.0 H Small | Kimi K2.5 |
|---|---|---|
| LMArena Instruction Following | 1221 | 1431 |
| IFEval | 89% | — |
Long Context Kimi K2.5 leads
Granite 4.0 H Small: 37.6 (#209), Kimi K2.5: 52.1 (#7)
| Benchmark | Granite 4.0 H Small | Kimi K2.5 |
|---|---|---|
| LMArena Longer Query | 1240 | 1445 |
| Fiction.LiveBench | — | 86.1% |
| CL-bench | — | 19.3% |
| CL-bench Life | — | 13.2% |
Writing & Preference Kimi K2.5 leads
Granite 4.0 H Small: 44.0 (#222), Kimi K2.5: 65.3 (#51)
| Benchmark | Granite 4.0 H Small | Kimi K2.5 |
|---|---|---|
| LMArena Text | 1241 | 1445 |
| LMArena Creative Writing | 1210 | 1423 |
| LMArena Multi-Turn | 1243 | 1444 |
| EQ-Bench Creative Writing | — | 1579 |
| WildBench | 73.9% | — |
Frequently asked questions
Is Granite 4.0 H Small better than Kimi K2.5?
Kimi K2.5 is the stronger model overall, scoring 47.8 to 35.9 on the Noometry Index.
Is Granite 4.0 H Small or Kimi K2.5 better for coding?
Kimi K2.5 scores higher on coding benchmarks: 48.7 versus 35.8 in the Noometry coding category.
How many benchmarks do Granite 4.0 H Small and Kimi K2.5 share?
14 benchmarks have published results for both models. Granite 4.0 H Small has 19 scored results on Noometry and Kimi K2.5 has 51.