Model comparison
GPT-4.1 vs Kimi K2 (Jul 2025)
Kimi K2 (Jul 2025) is the stronger model overall, scoring 41.2 to 35.9 on the Noometry Index.
Last verified . 36 shared benchmarks.
Summary
- They share 36 benchmarks with published results for both. GPT-4.1 scores higher in 2 categories and Kimi K2 (Jul 2025) in 7 categories; 6 gaps are clear of the uncertainty.
- The widest gap is in math, where Kimi K2 (Jul 2025) leads 42.7 to 22.3.
- The biggest single-benchmark swing is SWE-bench Verified (bash only): 39.6% for GPT-4.1 and 63.4% for Kimi K2 (Jul 2025).
- Kimi K2 (Jul 2025) is cheaper at $0.57 / $2.30 per million input/output tokens, against $2 / $8 for GPT-4.1.
- GPT-4.1 accepts more context: 1.05M tokens versus 262K.
- Kimi K2 (Jul 2025) has downloadable open weights; the other is API-only.
Side by side
| GPT-4.1 | Kimi K2 (Jul 2025) | |
|---|---|---|
| Provider | OpenAI | Moonshot AI |
| Noometry Index | 35.9 | 41.2 |
| Released | 2025-04-14 | 2025-07-12 |
| Weights | Proprietary | Open |
| Context window | 1.05M | 262K |
| Max output | 33K | 262K |
| Input $ / M tokens | $2 | $0.57 |
| Output $ / M tokens | $8 | $2.30 |
| Results tracked | 52 | 42 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Kimi K2 (Jul 2025) leads
GPT-4.1: 34.4 (#238), Kimi K2 (Jul 2025): 42.4 (#102)
| Benchmark | GPT-4.1 | Kimi K2 (Jul 2025) |
|---|---|---|
| SWE-bench Verified (bash only) | 39.6% | 63.4% |
| Aider Polyglot | 52.4% | 59.1% |
| WeirdML | 39% | 42.8% |
| LMArena Coding | 1391 | 1399 |
| ALE-Bench | 558.1 | 597.5 |
| SWE-bench Verified | 48.5% | — |
| GSO | — | 4.9% |
| CadEval | 42% | — |
Agentic & Tool Use GPT-4.1 leads
GPT-4.1: 34.7 (#43), Kimi K2 (Jul 2025): 32.4 (#64)
| Benchmark | GPT-4.1 | Kimi K2 (Jul 2025) |
|---|---|---|
| Berkeley Function Calling Leaderboard | 54% | 59.1% |
| Terminal-Bench | — | 35.7% |
| METR Time Horizons | — | 59.2% |
Reasoning Kimi K2 (Jul 2025) leads
GPT-4.1: 11.7 (#339), Kimi K2 (Jul 2025): 23.3 (#179)
| Benchmark | GPT-4.1 | Kimi K2 (Jul 2025) |
|---|---|---|
| SimpleBench | 27% | 26.3% |
| Kagi LLM Benchmark | 52.3% | 64.4% |
| LMArena Hard Prompts | 1384 | 1384 |
| Epoch Capabilities Index | 136.78 | 146.01 |
| ForecastBench | 61.5 | 60.2 |
| ARC-AGI-2 | 0.4% | — |
| ARC-AGI-1 | 5.5% | — |
| Chess Puzzles | 6% | — |
| EnigmaEval | 2.2% | — |
| DTBench | 68.3% | — |
| LMCA | 25.6% | — |
Math Kimi K2 (Jul 2025) leads
GPT-4.1: 22.3 (#280), Kimi K2 (Jul 2025): 42.7 (#83)
| Benchmark | GPT-4.1 | Kimi K2 (Jul 2025) |
|---|---|---|
| Omni-MATH | 47.1% | 65.4% |
| LMArena Math | 1370 | 1397 |
| FrontierMath (Feb 2025 set) | 5.5% | 21.4% |
| FrontierMath Tier 4 (v1) | 0% | 0% |
| FrontierMath (Tiers 1-3) | 6% | — |
| OTIS Mock AIME 2024-2025 | 38.3% | — |
| MATH Level 5 | 83% | — |
Knowledge Too close to call
GPT-4.1: 37.1 (#160), Kimi K2 (Jul 2025): 37.3 (#157)
| Benchmark | GPT-4.1 | Kimi K2 (Jul 2025) |
|---|---|---|
| MMLU-Pro | 81.1% | 81.9% |
| Vectara Hallucination Rate | 5.6% | 17.9% |
| GPQA (HELM) | 65.9% | 65.3% |
| LMArena Expert | 1364 | 1365 |
| GPQA Diamond | 66.9% | — |
| Humanity's Last Exam | 5.4% | — |
| SimpleQA Verified | 31.1% | — |
| Confabulations | — | 20.4% |
Multimodal Not comparable
GPT-4.1: 38.2 (#67), Kimi K2 (Jul 2025): —
| Benchmark | GPT-4.1 | Kimi K2 (Jul 2025) |
|---|---|---|
| LMArena Vision | 1211 | — |
| GeoBench | 72% | — |
Multilingual Too close to call
GPT-4.1: 49.4 (#133), Kimi K2 (Jul 2025): 49.6 (#130)
| Benchmark | GPT-4.1 | Kimi K2 (Jul 2025) |
|---|---|---|
| LMArena Non-English | 1370 | 1372 |
| LMArena Chinese | 1382 | 1415 |
| LMArena French | 1382 | 1379 |
| LMArena German | 1381 | 1387 |
| LMArena Japanese | 1319 | 1349 |
| LMArena Korean | 1339 | 1325 |
| LMArena Russian | 1377 | 1385 |
| LMArena Spanish | 1376 | 1386 |
Instruction Following Too close to call
GPT-4.1: 71.3 (#153), Kimi K2 (Jul 2025): 71.1 (#156)
| Benchmark | GPT-4.1 | Kimi K2 (Jul 2025) |
|---|---|---|
| IFEval | 83.8% | 85% |
| LMArena Instruction Following | 1367 | 1348 |
Long Context Kimi K2 (Jul 2025) leads
GPT-4.1: 40.0 (#163), Kimi K2 (Jul 2025): 41.2 (#145)
| Benchmark | GPT-4.1 | Kimi K2 (Jul 2025) |
|---|---|---|
| Fiction.LiveBench | 63.9% | 66.7% |
| LMArena Longer Query | 1385 | 1353 |
| CL-bench | — | 17.6% |
Writing & Preference Kimi K2 (Jul 2025) leads
GPT-4.1: 57.6 (#125), Kimi K2 (Jul 2025): 62.3 (#78)
| Benchmark | GPT-4.1 | Kimi K2 (Jul 2025) |
|---|---|---|
| LMArena Text | 1383 | 1380 |
| LMArena Creative Writing | 1363 | 1350 |
| EQ-Bench Creative Writing | 1420 | 1666 |
| WildBench | 85.4% | 86.2% |
| LMArena Multi-Turn | 1398 | 1371 |
| Short-Story Creative Writing | — | 85.6% |
Frequently asked questions
Is GPT-4.1 better than Kimi K2 (Jul 2025)?
Kimi K2 (Jul 2025) is the stronger model overall, scoring 41.2 to 35.9 on the Noometry Index.
Which is cheaper, GPT-4.1 or Kimi K2 (Jul 2025)?
Kimi K2 (Jul 2025) is cheaper. It lists at $0.57 per million input tokens and $2.30 per million output tokens; GPT-4.1 lists at $2 and $8.
Is GPT-4.1 or Kimi K2 (Jul 2025) better for coding?
Kimi K2 (Jul 2025) scores higher on coding benchmarks: 42.4 versus 34.4 in the Noometry coding category.
Which has the bigger context window?
GPT-4.1 does, with 1.05M tokens against 262K.
How many benchmarks do GPT-4.1 and Kimi K2 (Jul 2025) share?
36 benchmarks have published results for both models. GPT-4.1 has 52 scored results on Noometry and Kimi K2 (Jul 2025) has 42.