Model comparison
GPT-5 vs Kimi K3
Kimi K3 is the stronger model overall, scoring 59.5 to 50.9 on the Noometry Index. GPT-5 costs 1.7× less per token, which makes it the better buy when Kimi K3's lead doesn't matter for your workload.
Last verified . 38 shared benchmarks.
Summary
- They share 38 benchmarks with published results for both. GPT-5 scores higher in 2 categories and Kimi K3 in 8 categories; 10 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Kimi K3 leads 63.0 to 38.3.
- The biggest single-benchmark swing is ProofBench: 18% for GPT-5 and 87% for Kimi K3.
- GPT-5 is cheaper at $1.25 / $10 per million input/output tokens, against $3 / $15 for Kimi K3.
- Kimi K3 accepts more context: 1.05M tokens versus 400K.
- Kimi K3 has downloadable open weights; the other is API-only.
Side by side
| GPT-5 | Kimi K3 | |
|---|---|---|
| Provider | OpenAI | Moonshot AI |
| Noometry Index | 50.9 | 59.5 |
| Released | 2025-08-07 | 2026-07-16 |
| Weights | Proprietary | Open |
| Context window | 400K | 1.05M |
| Max output | 128K | 1.05M |
| Input $ / M tokens | $1.25 | $3 |
| Output $ / M tokens | $10 | $15 |
| Results tracked | 69 | 53 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Kimi K3 leads
GPT-5: 50.3 (#47), Kimi K3: 61.0 (#10)
| Benchmark | GPT-5 | Kimi K3 |
|---|---|---|
| LMArena WebDev | 1418 | 1654 |
| SciCode | 42.9% | 59.5% |
| WeirdML | 60.7% | 82.6% |
| LMArena Coding | 1436 | 1508 |
| ALE-Bench | 1,162 | 1,524 |
| SWE-bench Verified | 73.6% | — |
| DeepSWE | — | 68.5% |
| FrontierCode | — | 44.2% |
| SWE-bench Verified (bash only) | 65% | — |
| Aider Polyglot | 88% | — |
| FrontierSWE | — | 25.9% |
| GSO | 6.9% | — |
| AlgoTune | 1.67 | — |
Agentic & Tool Use Kimi K3 leads
GPT-5: 33.1 (#56), Kimi K3: 41.8 (#20)
| Benchmark | GPT-5 | Kimi K3 |
|---|---|---|
| Terminal-Bench | 49.6% | — |
| APEX-Agents | — | 50.6% |
| GDPval | 34.8% | — |
| Remote Labor Index | 1.7% | — |
| τ²-bench Banking | — | 37.1% |
| DeepResearch Bench | 49.6% | — |
| PostTrainBench | — | 32% |
| BALROG | 32.8% | — |
| GBAEval | — | 48.3% |
| GDP.pdf | — | 19% |
| LMArena Search | 1133 | — |
| METR Time Horizons | 69.6% | — |
| Vending-Bench 2 | — | 5,165 |
Reasoning Kimi K3 leads
GPT-5: 38.3 (#64), Kimi K3: 63.0 (#17)
| Benchmark | GPT-5 | Kimi K3 |
|---|---|---|
| ARC-AGI-2 | 9.9% | 60.4% |
| SimpleBench | 56.7% | 60.7% |
| ARC-AGI-1 | 65.7% | 94.5% |
| CritPt | 12.6% | 23.4% |
| Chess Puzzles | 37% | 39% |
| LMArena Hard Prompts | 1416 | 1496 |
| Mystery Game Puzzles | 23% | 26% |
| DTBench | 90.7% | 91.2% |
| LMCA | 40% | 52.7% |
| Epoch Capabilities Index | 150 | 157.45 |
| ForecastBench | 61.4 | 61.1 |
| Kagi LLM Benchmark | 72.7% | — |
| NYT Connections (extended) | — | 93.6% |
| EnigmaEval | 10.5% | — |
| EBR-Bench | 12.7% | — |
| Surface Evolver Bench | — | 95% |
Math Kimi K3 leads
GPT-5: 55.0 (#44), Kimi K3: 74.2 (#16)
| Benchmark | GPT-5 | Kimi K3 |
|---|---|---|
| FrontierMath (Tiers 1-3) | 55.4% | 72.2% |
| FrontierMath Tier 4 | 22% | 39% |
| OTIS Mock AIME 2024-2025 | 91.4% | 97.2% |
| ProofBench | 18% | 87% |
| LMArena Math | 1407 | 1491 |
| MathArena Final-Answer Competitions | — | 87.8% |
| Omni-MATH | 64.7% | — |
| MATH Level 5 | 98.1% | — |
| FrontierMath (Feb 2025 set) | 32.4% | — |
| FrontierMath Tier 4 (v1) | 12.5% | — |
Knowledge Kimi K3 leads
GPT-5: 56.6 (#43), Kimi K3: 63.2 (#21)
| Benchmark | GPT-5 | Kimi K3 |
|---|---|---|
| GPQA Diamond | 86.2% | 93.1% |
| SimpleQA Verified | 50.1% | 50.6% |
| LMArena Expert | 1419 | 1521 |
| Humanity's Last Exam | 25.3% | — |
| MMLU-Pro | 86.3% | — |
| Confabulations | 10.3% | — |
| Vectara Hallucination Rate | 14.7% | — |
| GPQA (HELM) | 79.2% | — |
Multimodal GPT-5 leads
GPT-5: 46.8 (#13), Kimi K3: 37.8 (#70)
| Benchmark | GPT-5 | Kimi K3 |
|---|---|---|
| LMArena Vision | 1232 | — |
| GeoBench | 81% | — |
| VPCT | 66% | — |
| Blueprint-Bench 2 | — | 29.5% |
| Furniture Assembly | — | 34.2% |
Multilingual Kimi K3 leads
GPT-5: 51.4 (#110), Kimi K3: 56.3 (#21)
| Benchmark | GPT-5 | Kimi K3 |
|---|---|---|
| LMArena Non-English | 1397 | 1466 |
| LMArena Chinese | 1422 | 1529 |
| LMArena French | 1410 | 1491 |
| LMArena German | 1416 | 1488 |
| LMArena Japanese | 1409 | 1487 |
| LMArena Korean | 1360 | 1458 |
| LMArena Russian | 1406 | 1482 |
| LMArena Spanish | 1399 | 1472 |
Instruction Following Kimi K3 leads
GPT-5: 73.8 (#113), Kimi K3: 77.7 (#14)
| Benchmark | GPT-5 | Kimi K3 |
|---|---|---|
| LMArena Instruction Following | 1388 | 1483 |
| IFEval | 87.5% | — |
Long Context GPT-5 leads
GPT-5: 69.5 (#2), Kimi K3: 45.8 (#29)
| Benchmark | GPT-5 | Kimi K3 |
|---|---|---|
| LMArena Longer Query | 1399 | 1494 |
| Fiction.LiveBench | 97.2% | — |
Writing & Preference Kimi K3 leads
GPT-5: 63.4 (#65), Kimi K3: 76.6 (#4)
| Benchmark | GPT-5 | Kimi K3 |
|---|---|---|
| LMArena Text | 1406 | 1476 |
| LMArena Creative Writing | 1365 | 1454 |
| EQ-Bench Creative Writing | 1627 | 2082 |
| LMArena Multi-Turn | 1426 | 1488 |
| Short-Story Creative Writing | 86% | — |
| WildBench | 85.7% | — |
| EQ-Bench 4 | — | 1339 |
Frequently asked questions
Is GPT-5 better than Kimi K3?
Kimi K3 is the stronger model overall, scoring 59.5 to 50.9 on the Noometry Index. GPT-5 costs 1.7× less per token, which makes it the better buy when Kimi K3's lead doesn't matter for your workload.
Which is cheaper, GPT-5 or Kimi K3?
GPT-5 is cheaper. It lists at $1.25 per million input tokens and $10 per million output tokens; Kimi K3 lists at $3 and $15.
Is GPT-5 or Kimi K3 better for coding?
Kimi K3 scores higher on coding benchmarks: 61.0 versus 50.3 in the Noometry coding category.
Which has the bigger context window?
Kimi K3 does, with 1.05M tokens against 400K.
How many benchmarks do GPT-5 and Kimi K3 share?
38 benchmarks have published results for both models. GPT-5 has 69 scored results on Noometry and Kimi K3 has 53.