Model comparison
GPT-4.1 vs Kimi K3
Kimi K3 is the stronger model overall, scoring 59.5 to 35.9 on the Noometry Index. GPT-4.1 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 . 32 shared benchmarks.
Summary
- They share 32 benchmarks with published results for both. GPT-4.1 scores higher in 1 category and Kimi K3 in 9 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in math, where Kimi K3 leads 74.2 to 22.3.
- The biggest single-benchmark swing is ARC-AGI-1: 5.5% for GPT-4.1 and 94.5% for Kimi K3.
- GPT-4.1 is cheaper at $2 / $8 per million input/output tokens, against $3 / $15 for Kimi K3.
- Kimi K3 accepts more context: 1.05M tokens versus 1.05M.
- Kimi K3 has downloadable open weights; the other is API-only.
Side by side
| GPT-4.1 | Kimi K3 | |
|---|---|---|
| Provider | OpenAI | Moonshot AI |
| Noometry Index | 35.9 | 59.5 |
| Released | 2025-04-14 | 2026-07-16 |
| Weights | Proprietary | Open |
| Context window | 1.05M | 1.05M |
| Max output | 33K | 1.05M |
| Input $ / M tokens | $2 | $3 |
| Output $ / M tokens | $8 | $15 |
| Results tracked | 52 | 53 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Kimi K3 leads
GPT-4.1: 34.4 (#238), Kimi K3: 61.0 (#10)
| Benchmark | GPT-4.1 | Kimi K3 |
|---|---|---|
| WeirdML | 39% | 82.6% |
| LMArena Coding | 1391 | 1508 |
| ALE-Bench | 558.1 | 1,524 |
| SWE-bench Verified | 48.5% | — |
| DeepSWE | — | 68.5% |
| FrontierCode | — | 44.2% |
| SWE-bench Verified (bash only) | 39.6% | — |
| Aider Polyglot | 52.4% | — |
| LMArena WebDev | — | 1654 |
| FrontierSWE | — | 25.9% |
| SciCode | — | 59.5% |
| CadEval | 42% | — |
Agentic & Tool Use Kimi K3 leads
GPT-4.1: 34.7 (#43), Kimi K3: 41.8 (#20)
| Benchmark | GPT-4.1 | Kimi K3 |
|---|---|---|
| APEX-Agents | — | 50.6% |
| Berkeley Function Calling Leaderboard | 54% | — |
| τ²-bench Banking | — | 37.1% |
| PostTrainBench | — | 32% |
| GBAEval | — | 48.3% |
| GDP.pdf | — | 19% |
| Vending-Bench 2 | — | 5,165 |
Reasoning Kimi K3 leads
GPT-4.1: 11.7 (#339), Kimi K3: 63.0 (#17)
| Benchmark | GPT-4.1 | Kimi K3 |
|---|---|---|
| ARC-AGI-2 | 0.4% | 60.4% |
| SimpleBench | 27% | 60.7% |
| ARC-AGI-1 | 5.5% | 94.5% |
| Chess Puzzles | 6% | 39% |
| LMArena Hard Prompts | 1384 | 1496 |
| DTBench | 68.3% | 91.2% |
| LMCA | 25.6% | 52.7% |
| Epoch Capabilities Index | 136.78 | 157.45 |
| ForecastBench | 61.5 | 61.1 |
| Kagi LLM Benchmark | 52.3% | — |
| NYT Connections (extended) | — | 93.6% |
| CritPt | — | 23.4% |
| EnigmaEval | 2.2% | — |
| Mystery Game Puzzles | — | 26% |
| Surface Evolver Bench | — | 95% |
Math Kimi K3 leads
GPT-4.1: 22.3 (#280), Kimi K3: 74.2 (#16)
| Benchmark | GPT-4.1 | Kimi K3 |
|---|---|---|
| FrontierMath (Tiers 1-3) | 6% | 72.2% |
| OTIS Mock AIME 2024-2025 | 38.3% | 97.2% |
| LMArena Math | 1370 | 1491 |
| FrontierMath Tier 4 | — | 39% |
| MathArena Final-Answer Competitions | — | 87.8% |
| ProofBench | — | 87% |
| Omni-MATH | 47.1% | — |
| MATH Level 5 | 83% | — |
| FrontierMath (Feb 2025 set) | 5.5% | — |
| FrontierMath Tier 4 (v1) | 0% | — |
Knowledge Kimi K3 leads
GPT-4.1: 37.1 (#160), Kimi K3: 63.2 (#21)
| Benchmark | GPT-4.1 | Kimi K3 |
|---|---|---|
| GPQA Diamond | 66.9% | 93.1% |
| SimpleQA Verified | 31.1% | 50.6% |
| LMArena Expert | 1364 | 1521 |
| Humanity's Last Exam | 5.4% | — |
| MMLU-Pro | 81.1% | — |
| Vectara Hallucination Rate | 5.6% | — |
| GPQA (HELM) | 65.9% | — |
Multimodal Too close to call
GPT-4.1: 38.2 (#67), Kimi K3: 37.8 (#70)
| Benchmark | GPT-4.1 | Kimi K3 |
|---|---|---|
| LMArena Vision | 1211 | — |
| GeoBench | 72% | — |
| Blueprint-Bench 2 | — | 29.5% |
| Furniture Assembly | — | 34.2% |
Multilingual Kimi K3 leads
GPT-4.1: 49.4 (#133), Kimi K3: 56.3 (#21)
| Benchmark | GPT-4.1 | Kimi K3 |
|---|---|---|
| LMArena Non-English | 1370 | 1466 |
| LMArena Chinese | 1382 | 1529 |
| LMArena French | 1382 | 1491 |
| LMArena German | 1381 | 1488 |
| LMArena Japanese | 1319 | 1487 |
| LMArena Korean | 1339 | 1458 |
| LMArena Russian | 1377 | 1482 |
| LMArena Spanish | 1376 | 1472 |
Instruction Following Kimi K3 leads
GPT-4.1: 71.3 (#153), Kimi K3: 77.7 (#14)
| Benchmark | GPT-4.1 | Kimi K3 |
|---|---|---|
| LMArena Instruction Following | 1367 | 1483 |
| IFEval | 83.8% | — |
Long Context Kimi K3 leads
GPT-4.1: 40.0 (#163), Kimi K3: 45.8 (#29)
| Benchmark | GPT-4.1 | Kimi K3 |
|---|---|---|
| LMArena Longer Query | 1385 | 1494 |
| Fiction.LiveBench | 63.9% | — |
Writing & Preference Kimi K3 leads
GPT-4.1: 57.6 (#125), Kimi K3: 76.6 (#4)
| Benchmark | GPT-4.1 | Kimi K3 |
|---|---|---|
| LMArena Text | 1383 | 1476 |
| LMArena Creative Writing | 1363 | 1454 |
| EQ-Bench Creative Writing | 1420 | 2082 |
| LMArena Multi-Turn | 1398 | 1488 |
| WildBench | 85.4% | — |
| EQ-Bench 4 | — | 1339 |
Frequently asked questions
Is GPT-4.1 better than Kimi K3?
Kimi K3 is the stronger model overall, scoring 59.5 to 35.9 on the Noometry Index. GPT-4.1 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-4.1 or Kimi K3?
GPT-4.1 is cheaper. It lists at $2 per million input tokens and $8 per million output tokens; Kimi K3 lists at $3 and $15.
Is GPT-4.1 or Kimi K3 better for coding?
Kimi K3 scores higher on coding benchmarks: 61.0 versus 34.4 in the Noometry coding category.
Which has the bigger context window?
Kimi K3 does, with 1.05M tokens against 1.05M.
How many benchmarks do GPT-4.1 and Kimi K3 share?
32 benchmarks have published results for both models. GPT-4.1 has 52 scored results on Noometry and Kimi K3 has 53.