Model comparison
Kimi K2.5 vs Qwen3-Coder 480B-A35B Instruct
Kimi K2.5 is the stronger model overall, scoring 47.8 to 37.7 on the Noometry Index.
Last verified . 23 shared benchmarks.
Alibaba (Qwen)Summary
- They share 23 benchmarks with published results for both. Kimi K2.5 scores higher in 9 categories and Qwen3-Coder 480B-A35B Instruct in 0 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where Kimi K2.5 leads 53.6 to 36.4.
- The biggest single-benchmark swing is Kagi LLM Benchmark: 78.5% for Kimi K2.5 and 49.5% for Qwen3-Coder 480B-A35B Instruct.
- Kimi K2.5 is cheaper at $0.45 / $2.25 per million input/output tokens, against $1.50 / $7.50 for Qwen3-Coder 480B-A35B Instruct.
Side by side
| Kimi K2.5 | Qwen3-Coder 480B-A35B Instruct | |
|---|---|---|
| Provider | Moonshot AI | Alibaba (Qwen) |
| Noometry Index | 47.8 | 37.7 |
| Released | 2026-01-27 | 2025-04 |
| Weights | Open | Open |
| Context window | 262K | 262K |
| Max output | 262K | 66K |
| Input $ / M tokens | $0.45 | $1.50 |
| Output $ / M tokens | $2.25 | $7.50 |
| Results tracked | 51 | 25 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Kimi K2.5 leads
Kimi K2.5: 48.7 (#48), Qwen3-Coder 480B-A35B Instruct: 36.2 (#199)
| Benchmark | Kimi K2.5 | Qwen3-Coder 480B-A35B Instruct |
|---|---|---|
| SWE-bench Verified (bash only) | 70.8% | 55.4% |
| LMArena WebDev | 1436 | 1275 |
| WeirdML | 45.6% | 41.2% |
| LMArena Coding | 1474 | 1413 |
| ALE-Bench | 821.65 | 461.45 |
| SWE-bench Verified | 73.8% | — |
| SWE-bench Multilingual | 67.3% | — |
| SciCode | 49% | — |
| GSO | — | 4.9% |
| AlgoTune | — | 1.44 |
Agentic & Tool Use Kimi K2.5 leads
Kimi K2.5: 34.0 (#44), Qwen3-Coder 480B-A35B Instruct: 23.7 (#116)
| Benchmark | Kimi K2.5 | Qwen3-Coder 480B-A35B Instruct |
|---|---|---|
| Terminal-Bench | 43.2% | 27.2% |
| OSWorld | 63.3% | — |
| Vending-Bench 2 | 1,198 | — |
Reasoning Kimi K2.5 leads
Kimi K2.5: 31.2 (#74), Qwen3-Coder 480B-A35B Instruct: 24.7 (#141)
| Benchmark | Kimi K2.5 | Qwen3-Coder 480B-A35B Instruct |
|---|---|---|
| Kagi LLM Benchmark | 78.5% | 49.5% |
| LMArena Hard Prompts | 1454 | 1372 |
| ARC-AGI-2 | 11.8% | — |
| SimpleBench | 46.8% | — |
| 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
Kimi K2.5: 50.3 (#46), Qwen3-Coder 480B-A35B Instruct: 34.9 (#151)
| Benchmark | Kimi K2.5 | Qwen3-Coder 480B-A35B Instruct |
|---|---|---|
| LMArena Math | 1470 | 1364 |
| MathArena Final-Answer Competitions | 62.3% | — |
| OTIS Mock AIME 2024-2025 | 92.2% | — |
| FrontierMath (Feb 2025 set) | 27.9% | — |
| FrontierMath Tier 4 (v1) | 4.2% | — |
Knowledge Kimi K2.5 leads
Kimi K2.5: 53.6 (#52), Qwen3-Coder 480B-A35B Instruct: 36.4 (#159)
| Benchmark | Kimi K2.5 | Qwen3-Coder 480B-A35B Instruct |
|---|---|---|
| LMArena Expert | 1466 | 1337 |
| GPQA Diamond | 87.6% | — |
| Humanity's Last Exam | 24.4% | — |
| SimpleQA Verified | 34.3% | — |
| Vectara Hallucination Rate | 14.2% | — |
Multimodal Not comparable
Kimi K2.5: 40.4 (#37), Qwen3-Coder 480B-A35B Instruct: —
| Benchmark | Kimi K2.5 | Qwen3-Coder 480B-A35B Instruct |
|---|---|---|
| LMArena Vision | 1269 | — |
| LMArena Document | 1430 | — |
Multilingual Kimi K2.5 leads
Kimi K2.5: 54.1 (#50), Qwen3-Coder 480B-A35B Instruct: 47.8 (#143)
| Benchmark | Kimi K2.5 | Qwen3-Coder 480B-A35B Instruct |
|---|---|---|
| LMArena Non-English | 1433 | 1345 |
| LMArena Chinese | 1495 | 1356 |
| LMArena French | 1454 | 1394 |
| LMArena German | 1441 | 1324 |
| LMArena Japanese | 1422 | 1309 |
| LMArena Korean | 1412 | 1307 |
| LMArena Russian | 1436 | 1367 |
| LMArena Spanish | 1450 | 1360 |
Instruction Following Kimi K2.5 leads
Kimi K2.5: 75.5 (#60), Qwen3-Coder 480B-A35B Instruct: 71.8 (#142)
| Benchmark | Kimi K2.5 | Qwen3-Coder 480B-A35B Instruct |
|---|---|---|
| LMArena Instruction Following | 1431 | 1356 |
Long Context Kimi K2.5 leads
Kimi K2.5: 52.1 (#7), Qwen3-Coder 480B-A35B Instruct: 42.0 (#126)
| Benchmark | Kimi K2.5 | Qwen3-Coder 480B-A35B Instruct |
|---|---|---|
| LMArena Longer Query | 1445 | 1379 |
| Fiction.LiveBench | 86.1% | — |
| CL-bench | 19.3% | — |
| CL-bench Life | 13.2% | — |
Writing & Preference Kimi K2.5 leads
Kimi K2.5: 65.3 (#51), Qwen3-Coder 480B-A35B Instruct: 55.5 (#142)
| Benchmark | Kimi K2.5 | Qwen3-Coder 480B-A35B Instruct |
|---|---|---|
| LMArena Text | 1445 | 1356 |
| LMArena Creative Writing | 1423 | 1333 |
| LMArena Multi-Turn | 1444 | 1365 |
| EQ-Bench Creative Writing | 1579 | — |
Frequently asked questions
Is Kimi K2.5 better than Qwen3-Coder 480B-A35B Instruct?
Kimi K2.5 is the stronger model overall, scoring 47.8 to 37.7 on the Noometry Index.
Which is cheaper, Kimi K2.5 or Qwen3-Coder 480B-A35B Instruct?
Kimi K2.5 is cheaper. It lists at $0.45 per million input tokens and $2.25 per million output tokens; Qwen3-Coder 480B-A35B Instruct lists at $1.50 and $7.50.
Is Kimi K2.5 or Qwen3-Coder 480B-A35B Instruct better for coding?
Kimi K2.5 scores higher on coding benchmarks: 48.7 versus 36.2 in the Noometry coding category.
Which has the bigger context window?
Both accept 262K tokens.
How many benchmarks do Kimi K2.5 and Qwen3-Coder 480B-A35B Instruct share?
23 benchmarks have published results for both models. Kimi K2.5 has 51 scored results on Noometry and Qwen3-Coder 480B-A35B Instruct has 25.