Model comparison
GPT-6 Sol vs Kimi K2.5
GPT-6 Sol is the stronger model overall, scoring 61.8 to 48.1 on the Noometry Index. Kimi K2.5 costs 4.4× less per token, which makes it the better buy when GPT-6 Sol's lead doesn't matter for your workload.
Last verified . 32 shared benchmarks.
Summary
- They share 32 benchmarks with published results for both. GPT-6 Sol scores higher in 7 categories and Kimi K2.5 in 3 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where GPT-6 Sol leads 74.0 to 31.2.
- The biggest single-benchmark swing is ARC-AGI-2: 89.6% for GPT-6 Sol and 11.8% for Kimi K2.5.
- Kimi K2.5 is cheaper at $0.45 / $2.25 per million input/output tokens, against $2 / $10 for GPT-6 Sol.
- GPT-6 Sol accepts more context: 1.05M tokens versus 262K.
- Kimi K2.5 has downloadable open weights; the other is API-only.
Side by side
| GPT-6 Sol | Kimi K2.5 | |
|---|---|---|
| Provider | OpenAI | Moonshot AI |
| Noometry Index | 61.8 | 48.1 |
| Released | 2026-09-22 | 2026-01-27 |
| Weights | Proprietary | Open |
| Context window | 1.05M | 262K |
| Max output | 128K | 262K |
| Input $ / M tokens | $2 | $0.45 |
| Output $ / M tokens | $10 | $2.25 |
| Results tracked | 45 | 51 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-6 Sol leads
GPT-6 Sol: 60.1 (#11), Kimi K2.5: 48.8 (#53)
| Benchmark | GPT-6 Sol | Kimi K2.5 |
|---|---|---|
| LMArena WebDev | 1688 | 1437 |
| SciCode | 57.6% | 49% |
| LMArena Coding | 1447 | 1474 |
| ALE-Bench | 2,462 | 821.65 |
| SWE-bench Verified | — | 73.8% |
| DeepSWE | 68.8% | — |
| FrontierCode | 49.3% | — |
| SWE-bench Verified (bash only) | — | 70.8% |
| SWE-bench Multilingual | — | 67.3% |
| WeirdML | — | 45.6% |
Agentic & Tool Use GPT-6 Sol leads
GPT-6 Sol: 37.2 (#36), Kimi K2.5: 34.2 (#48)
| Benchmark | GPT-6 Sol | Kimi K2.5 |
|---|---|---|
| Vending-Bench 2 | 14,428 | 1,198 |
| Terminal-Bench | — | 43.2% |
| APEX-Agents | 54.3% | — |
| OSWorld | — | 63.3% |
| GDP.pdf | 26.4% | — |
Reasoning GPT-6 Sol leads
GPT-6 Sol: 74.0 (#9), Kimi K2.5: 31.2 (#80)
| Benchmark | GPT-6 Sol | Kimi K2.5 |
|---|---|---|
| ARC-AGI-2 | 89.6% | 11.8% |
| NYT Connections (extended) | 90.1% | 69.9% |
| ARC-AGI-1 | 95.5% | 65.3% |
| CritPt | 30.9% | 3.1% |
| LMArena Hard Prompts | 1418 | 1453 |
| Epoch Capabilities Index | 162.72 | 148.03 |
| SimpleBench | — | 46.8% |
| Kagi LLM Benchmark | — | 78.5% |
| Chess Puzzles | — | 12% |
| EnigmaEval | — | 3.4% |
| Thematic Generalization | — | 69.4% |
| EBR-Bench | 53.3% | — |
| Mystery Game Puzzles | 56% | — |
| DTBench | 97.3% | — |
| LMCA | 59.1% | — |
Math GPT-6 Sol leads
GPT-6 Sol: 87.2 (#7), Kimi K2.5: 51.8 (#53)
| Benchmark | GPT-6 Sol | Kimi K2.5 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 100% | 92.2% |
| LMArena Math | 1402 | 1470 |
| FrontierMath (Tiers 1-3) | 89.8% | — |
| FrontierMath Tier 4 | 90% | — |
| MathArena Final-Answer Competitions | — | 62.3% |
| ProofBench | 83% | — |
| FrontierMath (Feb 2025 set) | — | 27.9% |
| FrontierMath Tier 4 (v1) | — | 4.2% |
Knowledge GPT-6 Sol leads
GPT-6 Sol: 64.8 (#15), Kimi K2.5: 53.6 (#56)
| Benchmark | GPT-6 Sol | Kimi K2.5 |
|---|---|---|
| GPQA Diamond | 94.3% | 87.6% |
| SimpleQA Verified | 60.7% | 34.3% |
| Vectara Hallucination Rate | 6.5% | 14.2% |
| LMArena Expert | 1439 | 1466 |
| Humanity's Last Exam | — | 24.4% |
Multimodal GPT-6 Sol leads
GPT-6 Sol: 47.6 (#10), Kimi K2.5: 41.1 (#39)
| Benchmark | GPT-6 Sol | Kimi K2.5 |
|---|---|---|
| LMArena Vision | 1245 | 1269 |
| Blueprint-Bench 2 | 36.9% | — |
| Furniture Assembly | 58.3% | — |
| LMArena Document | — | 1430 |
Multilingual Kimi K2.5 leads
GPT-6 Sol: 50.5 (#118), Kimi K2.5: 53.9 (#53)
| Benchmark | GPT-6 Sol | Kimi K2.5 |
|---|---|---|
| LMArena Non-English | 1385 | 1433 |
| LMArena Chinese | 1405 | 1495 |
| LMArena French | 1410 | 1454 |
| LMArena German | 1390 | 1441 |
| LMArena Japanese | 1385 | 1421 |
| LMArena Korean | 1341 | 1410 |
| LMArena Russian | 1401 | 1435 |
| LMArena Spanish | 1384 | 1450 |
Instruction Following Too close to call
GPT-6 Sol: 74.5 (#94), Kimi K2.5: 75.3 (#64)
| Benchmark | GPT-6 Sol | Kimi K2.5 |
|---|---|---|
| LMArena Instruction Following | 1412 | 1431 |
Long Context Kimi K2.5 leads
GPT-6 Sol: 43.1 (#108), Kimi K2.5: 52.1 (#7)
| Benchmark | GPT-6 Sol | Kimi K2.5 |
|---|---|---|
| LMArena Longer Query | 1411 | 1445 |
| Fiction.LiveBench | — | 86.1% |
| CL-bench | — | 19.3% |
| CL-bench Life | — | 13.2% |
Writing & Preference GPT-6 Sol leads
GPT-6 Sol: 71.9 (#18), Kimi K2.5: 65.1 (#53)
| Benchmark | GPT-6 Sol | Kimi K2.5 |
|---|---|---|
| LMArena Text | 1395 | 1445 |
| LMArena Creative Writing | 1378 | 1423 |
| EQ-Bench Creative Writing | 2125 | 1579 |
| LMArena Multi-Turn | 1412 | 1444 |
Frequently asked questions
Is GPT-6 Sol better than Kimi K2.5?
GPT-6 Sol is the stronger model overall, scoring 61.8 to 48.1 on the Noometry Index. Kimi K2.5 costs 4.4× less per token, which makes it the better buy when GPT-6 Sol's lead doesn't matter for your workload.
Which is cheaper, GPT-6 Sol or Kimi K2.5?
Kimi K2.5 is cheaper. It lists at $0.45 per million input tokens and $2.25 per million output tokens; GPT-6 Sol lists at $2 and $10.
Is GPT-6 Sol or Kimi K2.5 better for coding?
GPT-6 Sol scores higher on coding benchmarks: 60.1 versus 48.8 in the Noometry coding category.
Which has the bigger context window?
GPT-6 Sol does, with 1.05M tokens against 262K.
How many benchmarks do GPT-6 Sol and Kimi K2.5 share?
32 benchmarks have published results for both models. GPT-6 Sol has 45 scored results on Noometry and Kimi K2.5 has 51.