Model comparison
GPT-5 vs Kimi K2.6
GPT-5 is the stronger model overall, scoring 50.9 to 47.7 on the Noometry Index. Kimi K2.6 costs 2.0× less per token, which makes it the better buy when GPT-5's lead doesn't matter for your workload.
Last verified . 40 shared benchmarks.
Summary
- They share 40 benchmarks with published results for both. GPT-5 scores higher in 4 categories and Kimi K2.6 in 6 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in long context, where GPT-5 leads 69.5 to 44.9.
- The biggest single-benchmark swing is SimpleQA Verified: 50.1% for GPT-5 and 34.9% for Kimi K2.6.
- Kimi K2.6 is cheaper at $0.95 / $4 per million input/output tokens, against $1.25 / $10 for GPT-5.
- GPT-5 accepts more context: 400K tokens versus 262K.
- Kimi K2.6 has downloadable open weights; the other is API-only.
Side by side
| GPT-5 | Kimi K2.6 | |
|---|---|---|
| Provider | OpenAI | Moonshot AI |
| Noometry Index | 50.9 | 47.7 |
| Released | 2025-08-07 | 2026-04-20 |
| Weights | Proprietary | Open |
| Context window | 400K | 262K |
| Max output | 128K | 262K |
| Input $ / M tokens | $1.25 | $0.95 |
| Output $ / M tokens | $10 | $4 |
| Results tracked | 69 | 51 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Too close to call
GPT-5: 50.3 (#47), Kimi K2.6: 50.7 (#43)
| Benchmark | GPT-5 | Kimi K2.6 |
|---|---|---|
| SWE-bench Verified | 73.6% | 76.7% |
| LMArena WebDev | 1418 | 1509 |
| SciCode | 42.9% | 53.5% |
| WeirdML | 60.7% | 55.9% |
| LMArena Coding | 1436 | 1488 |
| ALE-Bench | 1,162 | 1,093 |
| SWE-bench Verified (bash only) | 65% | — |
| Aider Polyglot | 88% | — |
| GSO | 6.9% | — |
| AlgoTune | 1.67 | — |
Agentic & Tool Use GPT-5 leads
GPT-5: 33.1 (#56), Kimi K2.6: 21.9 (#137)
| Benchmark | GPT-5 | Kimi K2.6 |
|---|---|---|
| Terminal-Bench | 49.6% | — |
| OSWorld 2.0 | — | 4.6% |
| GDPval | 34.8% | — |
| Remote Labor Index | 1.7% | — |
| DeepResearch Bench | 49.6% | — |
| BALROG | 32.8% | — |
| ExploitBench | — | 18.4% |
| GBAEval | — | 0.9% |
| GDP.pdf | — | 12% |
| LMArena Search | 1133 | — |
| METR Time Horizons | 69.6% | — |
| Vending-Bench 2 | — | 6,205 |
Reasoning Kimi K2.6 leads
GPT-5: 38.3 (#64), Kimi K2.6: 40.5 (#55)
| Benchmark | GPT-5 | Kimi K2.6 |
|---|---|---|
| CritPt | 12.6% | 8% |
| Chess Puzzles | 37% | 26% |
| EBR-Bench | 12.7% | 2.4% |
| LMArena Hard Prompts | 1416 | 1470 |
| Mystery Game Puzzles | 23% | 18% |
| DTBench | 90.7% | 90.9% |
| LMCA | 40% | 37.3% |
| Epoch Capabilities Index | 150 | 151.05 |
| ARC-AGI-2 | 9.9% | — |
| SimpleBench | 56.7% | — |
| Kagi LLM Benchmark | 72.7% | — |
| NYT Connections (extended) | — | 87.2% |
| ARC-AGI-1 | 65.7% | — |
| EnigmaEval | 10.5% | — |
| ForecastBench | 61.4 | — |
Math Kimi K2.6 leads
GPT-5: 55.0 (#44), Kimi K2.6: 57.0 (#41)
| Benchmark | GPT-5 | Kimi K2.6 |
|---|---|---|
| FrontierMath (Tiers 1-3) | 55.4% | 57.2% |
| FrontierMath Tier 4 | 22% | 25.6% |
| OTIS Mock AIME 2024-2025 | 91.4% | 96.1% |
| ProofBench | 18% | 16% |
| LMArena Math | 1407 | 1475 |
| FrontierMath (Feb 2025 set) | 32.4% | 39% |
| FrontierMath Tier 4 (v1) | 12.5% | 14.6% |
| MathArena Final-Answer Competitions | — | 72.9% |
| Omni-MATH | 64.7% | — |
| MATH Level 5 | 98.1% | — |
Knowledge GPT-5 leads
GPT-5: 56.6 (#43), Kimi K2.6: 54.0 (#54)
| Benchmark | GPT-5 | Kimi K2.6 |
|---|---|---|
| GPQA Diamond | 86.2% | 90.8% |
| SimpleQA Verified | 50.1% | 34.9% |
| Vectara Hallucination Rate | 14.7% | 10.8% |
| LMArena Expert | 1419 | 1491 |
| Humanity's Last Exam | 25.3% | — |
| MMLU-Pro | 86.3% | — |
| Confabulations | 10.3% | — |
| GPQA (HELM) | 79.2% | — |
Multimodal GPT-5 leads
GPT-5: 46.8 (#13), Kimi K2.6: 31.6 (#103)
| Benchmark | GPT-5 | Kimi K2.6 |
|---|---|---|
| LMArena Vision | 1232 | 1283 |
| GeoBench | 81% | — |
| VPCT | 66% | — |
| Blueprint-Bench 2 | — | 3.9% |
| Furniture Assembly | — | 21.7% |
| LMArena Document | — | 1451 |
Multilingual Kimi K2.6 leads
GPT-5: 51.4 (#110), Kimi K2.6: 54.9 (#37)
| Benchmark | GPT-5 | Kimi K2.6 |
|---|---|---|
| LMArena Non-English | 1397 | 1446 |
| LMArena Chinese | 1422 | 1521 |
| LMArena French | 1410 | 1471 |
| LMArena German | 1416 | 1450 |
| LMArena Japanese | 1409 | 1443 |
| LMArena Korean | 1360 | 1427 |
| LMArena Russian | 1406 | 1446 |
| LMArena Spanish | 1399 | 1464 |
Instruction Following Kimi K2.6 leads
GPT-5: 73.8 (#113), Kimi K2.6: 76.3 (#43)
| Benchmark | GPT-5 | Kimi K2.6 |
|---|---|---|
| LMArena Instruction Following | 1388 | 1451 |
| IFEval | 87.5% | — |
Long Context GPT-5 leads
GPT-5: 69.5 (#2), Kimi K2.6: 44.9 (#52)
| Benchmark | GPT-5 | Kimi K2.6 |
|---|---|---|
| LMArena Longer Query | 1399 | 1468 |
| Fiction.LiveBench | 97.2% | — |
Writing & Preference Kimi K2.6 leads
GPT-5: 63.4 (#65), Kimi K2.6: 68.5 (#26)
| Benchmark | GPT-5 | Kimi K2.6 |
|---|---|---|
| LMArena Text | 1406 | 1455 |
| LMArena Creative Writing | 1365 | 1434 |
| EQ-Bench Creative Writing | 1627 | 1725 |
| LMArena Multi-Turn | 1426 | 1453 |
| Short-Story Creative Writing | 86% | — |
| WildBench | 85.7% | — |
| EQ-Bench 4 | — | 1202 |
Frequently asked questions
Is GPT-5 better than Kimi K2.6?
GPT-5 is the stronger model overall, scoring 50.9 to 47.7 on the Noometry Index. Kimi K2.6 costs 2.0× less per token, which makes it the better buy when GPT-5's lead doesn't matter for your workload.
Which is cheaper, GPT-5 or Kimi K2.6?
Kimi K2.6 is cheaper. It lists at $0.95 per million input tokens and $4 per million output tokens; GPT-5 lists at $1.25 and $10.
Is GPT-5 or Kimi K2.6 better for coding?
They score almost the same on coding (50.3 vs 50.7); test both on your own repository before choosing.
Which has the bigger context window?
GPT-5 does, with 400K tokens against 262K.
How many benchmarks do GPT-5 and Kimi K2.6 share?
40 benchmarks have published results for both models. GPT-5 has 69 scored results on Noometry and Kimi K2.6 has 51.