Model comparison
GPT-5 Mini vs Kimi K2 Thinking Turbo
Kimi K2 Thinking Turbo is the stronger model overall, scoring 45.4 to 41.7 on the Noometry Index.
Last verified . 20 shared benchmarks.
Summary
- They share 20 benchmarks with published results for both. GPT-5 Mini scores higher in 3 categories and Kimi K2 Thinking Turbo in 5 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Kimi K2 Thinking Turbo leads 31.4 to 24.1.
- The biggest single-benchmark swing is Chess Puzzles: 30% for GPT-5 Mini and 20% for Kimi K2 Thinking Turbo.
- Kimi K2 Thinking Turbo has downloadable open weights; the other is API-only.
Side by side
| GPT-5 Mini | Kimi K2 Thinking Turbo | |
|---|---|---|
| Provider | OpenAI | Moonshot AI |
| Noometry Index | 41.7 | 45.4 |
| Released | 2025-08-07 | 2025-11-06 |
| Weights | Proprietary | Open |
| Context window | 400K | — |
| Max output | 128K | — |
| Input $ / M tokens | $0.25 | — |
| Output $ / M tokens | $2 | — |
| Results tracked | 60 | 21 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-5 Mini leads
GPT-5 Mini: 40.0 (#134), Kimi K2 Thinking Turbo: 38.1 (#166)
| Benchmark | GPT-5 Mini | Kimi K2 Thinking Turbo |
|---|---|---|
| LMArena Coding | 1406 | 1454 |
| SWE-bench Verified | 64.7% | — |
| SWE-bench Verified (bash only) | 59.8% | — |
| LMArena WebDev | — | 1322 |
| SWE-bench Multilingual | 39.7% | — |
| SciCode | 39.2% | — |
| WeirdML | 52.7% | — |
| ALE-Bench | 799.77 | — |
| AlgoTune | 1.38 | — |
Agentic & Tool Use Not comparable
GPT-5 Mini: 31.0 (#64), Kimi K2 Thinking Turbo: —
| Benchmark | GPT-5 Mini | Kimi K2 Thinking Turbo |
|---|---|---|
| Terminal-Bench | 34.8% | — |
| Berkeley Function Calling Leaderboard | 55.5% | — |
| Vending-Bench 2 | -31.18 | — |
Reasoning Kimi K2 Thinking Turbo leads
GPT-5 Mini: 24.1 (#153), Kimi K2 Thinking Turbo: 31.4 (#72)
| Benchmark | GPT-5 Mini | Kimi K2 Thinking Turbo |
|---|---|---|
| Chess Puzzles | 30% | 20% |
| LMArena Hard Prompts | 1380 | 1428 |
| ARC-AGI-2 | 4.4% | — |
| Kagi LLM Benchmark | 70.3% | — |
| ARC-AGI-1 | 54.3% | — |
| CritPt | 0% | — |
| EnigmaEval | 8.2% | — |
| Mystery Game Puzzles | 10% | — |
| DTBench | 80.5% | — |
| LMCA | 34.2% | — |
| Epoch Capabilities Index | 145.51 | — |
| ForecastBench | 61 | — |
Math Too close to call
GPT-5 Mini: 46.3 (#59), Kimi K2 Thinking Turbo: 45.5 (#62)
| Benchmark | GPT-5 Mini | Kimi K2 Thinking Turbo |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 86.7% | 83.1% |
| LMArena Math | 1377 | 1429 |
| FrontierMath (Tiers 1-3) | 46.7% | — |
| FrontierMath Tier 4 | 12.2% | — |
| ProofBench | 9% | — |
| Omni-MATH | 72.2% | — |
| MATH Level 5 | 97.8% | — |
| FrontierMath (Feb 2025 set) | 27.2% | — |
| FrontierMath Tier 4 (v1) | 6.3% | — |
Knowledge Kimi K2 Thinking Turbo leads
GPT-5 Mini: 45.5 (#80), Kimi K2 Thinking Turbo: 50.9 (#64)
| Benchmark | GPT-5 Mini | Kimi K2 Thinking Turbo |
|---|---|---|
| GPQA Diamond | 75% | 84.2% |
| LMArena Expert | 1377 | 1439 |
| Humanity's Last Exam | 19.4% | — |
| SimpleQA Verified | 21.6% | — |
| MMLU-Pro | 83.5% | — |
| Confabulations | 13.3% | — |
| Vectara Hallucination Rate | 12.9% | — |
| GPQA (HELM) | 75.6% | — |
Multimodal Not comparable
GPT-5 Mini: 35.0 (#80), Kimi K2 Thinking Turbo: —
| Benchmark | GPT-5 Mini | Kimi K2 Thinking Turbo |
|---|---|---|
| LMArena Vision | 1202 | — |
| VPCT | 40.2% | — |
Multilingual Kimi K2 Thinking Turbo leads
GPT-5 Mini: 49.1 (#131), Kimi K2 Thinking Turbo: 51.6 (#104)
| Benchmark | GPT-5 Mini | Kimi K2 Thinking Turbo |
|---|---|---|
| LMArena Non-English | 1363 | 1398 |
| LMArena Chinese | 1387 | 1457 |
| LMArena French | 1385 | 1425 |
| LMArena German | 1366 | 1390 |
| LMArena Japanese | 1342 | 1356 |
| LMArena Korean | 1306 | 1329 |
| LMArena Russian | 1362 | 1390 |
| LMArena Spanish | 1355 | 1407 |
Instruction Following GPT-5 Mini leads
GPT-5 Mini: 76.4 (#43), Kimi K2 Thinking Turbo: 74.2 (#103)
| Benchmark | GPT-5 Mini | Kimi K2 Thinking Turbo |
|---|---|---|
| LMArena Instruction Following | 1357 | 1403 |
| IFEval | 92.7% | — |
Long Context Kimi K2 Thinking Turbo leads
GPT-5 Mini: 41.9 (#127), Kimi K2 Thinking Turbo: 43.2 (#99)
| Benchmark | GPT-5 Mini | Kimi K2 Thinking Turbo |
|---|---|---|
| LMArena Longer Query | 1355 | 1415 |
| Fiction.LiveBench | 69.4% | — |
Writing & Preference Kimi K2 Thinking Turbo leads
GPT-5 Mini: 55.4 (#143), Kimi K2 Thinking Turbo: 60.2 (#99)
| Benchmark | GPT-5 Mini | Kimi K2 Thinking Turbo |
|---|---|---|
| LMArena Text | 1373 | 1415 |
| LMArena Creative Writing | 1325 | 1373 |
| LMArena Multi-Turn | 1363 | 1414 |
| Short-Story Creative Writing | 83.1% | — |
| EQ-Bench Creative Writing | 1313 | — |
| WildBench | 85.5% | — |
Frequently asked questions
Is GPT-5 Mini better than Kimi K2 Thinking Turbo?
Kimi K2 Thinking Turbo is the stronger model overall, scoring 45.4 to 41.7 on the Noometry Index.
Is GPT-5 Mini or Kimi K2 Thinking Turbo better for coding?
GPT-5 Mini scores higher on coding benchmarks: 40.0 versus 38.1 in the Noometry coding category.
How many benchmarks do GPT-5 Mini and Kimi K2 Thinking Turbo share?
20 benchmarks have published results for both models. GPT-5 Mini has 60 scored results on Noometry and Kimi K2 Thinking Turbo has 21.