Model comparison
Kimi K2.5 vs Nemotron 3 Ultra
Kimi K2.5 is the stronger model overall, scoring 48.1 to 42.5 on the Noometry Index.
Last verified . 24 shared benchmarks.
NVIDIASummary
- They share 24 benchmarks with published results for both. Kimi K2.5 scores higher in 8 categories and Nemotron 3 Ultra in 1 category; 6 gaps are clear of the uncertainty.
- The widest gap is in math, where Kimi K2.5 leads 51.8 to 35.3.
- The biggest single-benchmark swing is SciCode: 49% for Kimi K2.5 and 40.3% for Nemotron 3 Ultra.
- Kimi K2.5 is cheaper at $0.45 / $2.25 per million input/output tokens, against $0.60 / $2.40 for Nemotron 3 Ultra.
Side by side
| Kimi K2.5 | Nemotron 3 Ultra | |
|---|---|---|
| Provider | Moonshot AI | NVIDIA |
| Noometry Index | 48.1 | 42.5 |
| Released | 2026-01-27 | 2026-06-04 |
| Weights | Open | Open |
| Context window | 262K | 262K |
| Max output | 262K | 128K |
| Input $ / M tokens | $0.45 | $0.60 |
| Output $ / M tokens | $2.25 | $2.40 |
| Results tracked | 51 | 30 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Kimi K2.5 leads
Kimi K2.5: 48.8 (#53), Nemotron 3 Ultra: 38.1 (#182)
| Benchmark | Kimi K2.5 | Nemotron 3 Ultra |
|---|---|---|
| SciCode | 49% | 40.3% |
| WeirdML | 45.6% | 43.5% |
| LMArena Coding | 1474 | 1468 |
| SWE-bench Verified | 73.8% | — |
| FrontierCode | — | 13.6% |
| SWE-bench Verified (bash only) | 70.8% | — |
| LMArena WebDev | 1437 | — |
| SWE-bench Multilingual | 67.3% | — |
| ALE-Bench | 821.65 | — |
Agentic & Tool Use Kimi K2.5 leads
Kimi K2.5: 34.2 (#48), Nemotron 3 Ultra: 23.4 (#126)
| Benchmark | Kimi K2.5 | Nemotron 3 Ultra |
|---|---|---|
| Terminal-Bench | 43.2% | — |
| APEX-Agents | — | 22.7% |
| OSWorld | 63.3% | — |
| Vending-Bench 2 | 1,198 | — |
Reasoning Too close to call
Kimi K2.5: 31.2 (#80), Nemotron 3 Ultra: 31.0 (#82)
| Benchmark | Kimi K2.5 | Nemotron 3 Ultra |
|---|---|---|
| CritPt | 3.1% | 3.1% |
| Chess Puzzles | 12% | 12% |
| LMArena Hard Prompts | 1453 | 1452 |
| Epoch Capabilities Index | 148.03 | 146.17 |
| ARC-AGI-2 | 11.8% | — |
| SimpleBench | 46.8% | — |
| Kagi LLM Benchmark | 78.5% | — |
| NYT Connections (extended) | 69.9% | — |
| ARC-AGI-1 | 65.3% | — |
| EnigmaEval | 3.4% | — |
| Thematic Generalization | 69.4% | — |
| Mystery Game Puzzles | — | 20% |
| DTBench | — | 90.1% |
| LMCA | — | 36.9% |
Math Kimi K2.5 leads
Kimi K2.5: 51.8 (#53), Nemotron 3 Ultra: 35.3 (#187)
| Benchmark | Kimi K2.5 | Nemotron 3 Ultra |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 92.2% | 86.7% |
| LMArena Math | 1470 | 1457 |
| MathArena Final-Answer Competitions | 62.3% | — |
| ProofBench | — | 2% |
| FrontierMath (Feb 2025 set) | 27.9% | — |
| FrontierMath Tier 4 (v1) | 4.2% | — |
Knowledge Kimi K2.5 leads
Kimi K2.5: 53.6 (#56), Nemotron 3 Ultra: 52.5 (#61)
| Benchmark | Kimi K2.5 | Nemotron 3 Ultra |
|---|---|---|
| GPQA Diamond | 87.6% | 85.4% |
| LMArena Expert | 1466 | 1472 |
| Humanity's Last Exam | 24.4% | — |
| SimpleQA Verified | 34.3% | — |
| Vectara Hallucination Rate | 14.2% | — |
Multimodal Not comparable
Kimi K2.5: 41.1 (#39), Nemotron 3 Ultra: —
| Benchmark | Kimi K2.5 | Nemotron 3 Ultra |
|---|---|---|
| LMArena Vision | 1269 | — |
| LMArena Document | 1430 | — |
Multilingual Too close to call
Kimi K2.5: 53.9 (#53), Nemotron 3 Ultra: 53.5 (#64)
| Benchmark | Kimi K2.5 | Nemotron 3 Ultra |
|---|---|---|
| LMArena Non-English | 1433 | 1427 |
| LMArena Chinese | 1495 | 1497 |
| LMArena French | 1454 | 1472 |
| LMArena German | 1441 | 1471 |
| LMArena Korean | 1410 | 1386 |
| LMArena Russian | 1435 | 1417 |
| LMArena Spanish | 1450 | 1454 |
| LMArena Japanese | 1421 | — |
Instruction Following Too close to call
Kimi K2.5: 75.3 (#64), Nemotron 3 Ultra: 74.7 (#88)
| Benchmark | Kimi K2.5 | Nemotron 3 Ultra |
|---|---|---|
| LMArena Instruction Following | 1431 | 1418 |
Long Context Kimi K2.5 leads
Kimi K2.5: 52.1 (#7), Nemotron 3 Ultra: 43.9 (#81)
| Benchmark | Kimi K2.5 | Nemotron 3 Ultra |
|---|---|---|
| LMArena Longer Query | 1445 | 1437 |
| Fiction.LiveBench | 86.1% | — |
| CL-bench | 19.3% | — |
| CL-bench Life | 13.2% | — |
Writing & Preference Nemotron 3 Ultra leads
Kimi K2.5: 65.1 (#53), Nemotron 3 Ultra: 66.3 (#36)
| Benchmark | Kimi K2.5 | Nemotron 3 Ultra |
|---|---|---|
| LMArena Text | 1445 | 1445 |
| LMArena Creative Writing | 1423 | 1398 |
| EQ-Bench Creative Writing | 1579 | 1692 |
| LMArena Multi-Turn | 1444 | 1411 |
Frequently asked questions
Is Kimi K2.5 better than Nemotron 3 Ultra?
Kimi K2.5 is the stronger model overall, scoring 48.1 to 42.5 on the Noometry Index.
Which is cheaper, Kimi K2.5 or Nemotron 3 Ultra?
Kimi K2.5 is cheaper. It lists at $0.45 per million input tokens and $2.25 per million output tokens; Nemotron 3 Ultra lists at $0.60 and $2.40.
Is Kimi K2.5 or Nemotron 3 Ultra better for coding?
Kimi K2.5 scores higher on coding benchmarks: 48.8 versus 38.1 in the Noometry coding category.
Which has the bigger context window?
Both accept 262K tokens.
How many benchmarks do Kimi K2.5 and Nemotron 3 Ultra share?
24 benchmarks have published results for both models. Kimi K2.5 has 51 scored results on Noometry and Nemotron 3 Ultra has 30.