Model comparison
GPT-6 Luna vs Kimi K3
Kimi K3 is the stronger model overall, scoring 59.5 to 45.1 on the Noometry Index. GPT-6 Luna costs 30× less per token, which makes it the better buy when Kimi K3's lead doesn't matter for your workload.
Last verified . 18 shared benchmarks.
Summary
- They share 18 benchmarks with published results for both. GPT-6 Luna scores higher in 1 category and Kimi K3 in 8 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in math, where Kimi K3 leads 74.0 to 35.6.
- The biggest single-benchmark swing is NYT Connections (extended): 68.7% for GPT-6 Luna and 93.6% for Kimi K3.
- GPT-6 Luna is cheaper at $0.10 / $0.50 per million input/output tokens, against $3 / $15 for Kimi K3.
- GPT-6 Luna accepts more context: 1.05M tokens versus 1.05M.
- Kimi K3 has downloadable open weights; the other is API-only.
Side by side
| GPT-6 Luna | Kimi K3 | |
|---|---|---|
| Provider | OpenAI | Moonshot AI |
| Noometry Index | 45.1 | 59.5 |
| Released | 2026-09-22 | 2026-07-16 |
| Weights | Proprietary | Open |
| Context window | 1.05M | 1.05M |
| Max output | 128K | 1.05M |
| Input $ / M tokens | $0.10 | $3 |
| Output $ / M tokens | $0.50 | $15 |
| Results tracked | 18 | 53 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Kimi K3 leads
GPT-6 Luna: 54.7 (#24), Kimi K3: 60.6 (#8)
| Benchmark | GPT-6 Luna | Kimi K3 |
|---|---|---|
| DeepSWE | 66.6% | 68.5% |
| FrontierCode | 42.4% | 44.2% |
| LMArena WebDev | 1583 | 1659 |
| LMArena Coding | 1448 | 1509 |
| FrontierSWE | — | 25.9% |
| SciCode | — | 59.5% |
| WeirdML | — | 82.6% |
| ALE-Bench | — | 1,524 |
Agentic & Tool Use Not comparable
GPT-6 Luna: —, Kimi K3: 41.8 (#19)
| Benchmark | GPT-6 Luna | Kimi K3 |
|---|---|---|
| APEX-Agents | — | 50.6% |
| τ²-bench Banking | — | 37.1% |
| PostTrainBench | — | 32% |
| GBAEval | — | 48.3% |
| GDP.pdf | — | 19% |
| Vending-Bench 2 | — | 5,165 |
Reasoning Kimi K3 leads
GPT-6 Luna: 31.7 (#68), Kimi K3: 63.0 (#15)
| Benchmark | GPT-6 Luna | Kimi K3 |
|---|---|---|
| NYT Connections (extended) | 68.7% | 93.6% |
| LMArena Hard Prompts | 1415 | 1495 |
| ARC-AGI-2 | — | 60.4% |
| SimpleBench | — | 60.7% |
| ARC-AGI-1 | — | 94.5% |
| CritPt | — | 23.4% |
| Chess Puzzles | — | 39% |
| Mystery Game Puzzles | — | 26% |
| DTBench | — | 91.2% |
| LMCA | — | 52.7% |
| Surface Evolver Bench | — | 95% |
| Epoch Capabilities Index | — | 157.68 |
| ForecastBench | — | 61.1 |
Math Kimi K3 leads
GPT-6 Luna: 35.6 (#140), Kimi K3: 74.0 (#13)
| Benchmark | GPT-6 Luna | Kimi K3 |
|---|---|---|
| LMArena Math | 1396 | 1493 |
| FrontierMath (Tiers 1-3) | — | 72.2% |
| FrontierMath Tier 4 | — | 39% |
| MathArena Final-Answer Competitions | — | 87.8% |
| OTIS Mock AIME 2024-2025 | — | 97.2% |
| ProofBench | — | 87% |
Knowledge Kimi K3 leads
GPT-6 Luna: 40.1 (#110), Kimi K3: 63.2 (#19)
| Benchmark | GPT-6 Luna | Kimi K3 |
|---|---|---|
| LMArena Expert | 1458 | 1520 |
| GPQA Diamond | — | 93.1% |
| SimpleQA Verified | — | 50.6% |
Multimodal GPT-6 Luna leads
GPT-6 Luna: 42.0 (#27), Kimi K3: 38.4 (#58)
| Benchmark | GPT-6 Luna | Kimi K3 |
|---|---|---|
| Furniture Assembly | 44.2% | 34.2% |
| Blueprint-Bench 2 | — | 29.5% |
Multilingual Kimi K3 leads
GPT-6 Luna: 51.0 (#111), Kimi K3: 56.5 (#20)
| Benchmark | GPT-6 Luna | Kimi K3 |
|---|---|---|
| LMArena Non-English | 1390 | 1466 |
| LMArena Chinese | 1425 | 1528 |
| LMArena Russian | 1392 | 1481 |
| LMArena Spanish | 1416 | 1478 |
| LMArena French | — | 1488 |
| LMArena German | — | 1474 |
| LMArena Japanese | — | 1492 |
| LMArena Korean | — | 1465 |
Instruction Following Kimi K3 leads
GPT-6 Luna: 74.7 (#85), Kimi K3: 77.8 (#13)
| Benchmark | GPT-6 Luna | Kimi K3 |
|---|---|---|
| LMArena Instruction Following | 1414 | 1481 |
Long Context Kimi K3 leads
GPT-6 Luna: 43.3 (#95), Kimi K3: 45.8 (#27)
| Benchmark | GPT-6 Luna | Kimi K3 |
|---|---|---|
| LMArena Longer Query | 1417 | 1494 |
Writing & Preference Kimi K3 leads
GPT-6 Luna: 58.4 (#115), Kimi K3: 76.7 (#4)
| Benchmark | GPT-6 Luna | Kimi K3 |
|---|---|---|
| LMArena Text | 1392 | 1476 |
| LMArena Creative Writing | 1357 | 1452 |
| LMArena Multi-Turn | 1400 | 1488 |
| EQ-Bench Creative Writing | — | 2082 |
| EQ-Bench 4 | — | 1339 |
Frequently asked questions
Is GPT-6 Luna better than Kimi K3?
Kimi K3 is the stronger model overall, scoring 59.5 to 45.1 on the Noometry Index. GPT-6 Luna costs 30× less per token, which makes it the better buy when Kimi K3's lead doesn't matter for your workload.
Which is cheaper, GPT-6 Luna or Kimi K3?
GPT-6 Luna is cheaper. It lists at $0.10 per million input tokens and $0.50 per million output tokens; Kimi K3 lists at $3 and $15.
Is GPT-6 Luna or Kimi K3 better for coding?
Kimi K3 scores higher on coding benchmarks: 60.6 versus 54.7 in the Noometry coding category.
Which has the bigger context window?
GPT-6 Luna does, with 1.05M tokens against 1.05M.
How many benchmarks do GPT-6 Luna and Kimi K3 share?
18 benchmarks have published results for both models. GPT-6 Luna has 18 scored results on Noometry and Kimi K3 has 53.