Model comparison
GPT-5 Mini vs Kimi K3
Kimi K3 is the stronger model overall, scoring 59.5 to 41.8 on the Noometry Index. GPT-5 Mini costs 8.7× less per token, which makes it the better buy when Kimi K3's lead doesn't matter for your workload.
Last verified . 37 shared benchmarks.
Summary
- They share 37 benchmarks with published results for both. GPT-5 Mini scores higher in 0 categories and Kimi K3 in 10 categories; 10 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Kimi K3 leads 63.0 to 23.9.
- The biggest single-benchmark swing is ProofBench: 9% for GPT-5 Mini and 87% for Kimi K3.
- GPT-5 Mini is cheaper at $0.25 / $2 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 Mini | Kimi K3 | |
|---|---|---|
| Provider | OpenAI | Moonshot AI |
| Noometry Index | 41.8 | 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 | $0.25 | $3 |
| Output $ / M tokens | $2 | $15 |
| Results tracked | 60 | 53 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Kimi K3 leads
GPT-5 Mini: 40.1 (#146), Kimi K3: 61.0 (#10)
| Benchmark | GPT-5 Mini | Kimi K3 |
|---|---|---|
| SciCode | 39.2% | 59.5% |
| WeirdML | 52.7% | 82.6% |
| LMArena Coding | 1406 | 1508 |
| ALE-Bench | 799.77 | 1,524 |
| SWE-bench Verified | 64.7% | — |
| DeepSWE | — | 68.5% |
| FrontierCode | — | 44.2% |
| SWE-bench Verified (bash only) | 59.8% | — |
| LMArena WebDev | — | 1654 |
| SWE-bench Multilingual | 39.7% | — |
| FrontierSWE | — | 25.9% |
| AlgoTune | 1.38 | — |
Agentic & Tool Use Kimi K3 leads
GPT-5 Mini: 31.1 (#70), Kimi K3: 41.8 (#20)
| Benchmark | GPT-5 Mini | Kimi K3 |
|---|---|---|
| Vending-Bench 2 | -31.18 | 5,165 |
| Terminal-Bench | 34.8% | — |
| APEX-Agents | — | 50.6% |
| Berkeley Function Calling Leaderboard | 55.5% | — |
| τ²-bench Banking | — | 37.1% |
| PostTrainBench | — | 32% |
| GBAEval | — | 48.3% |
| GDP.pdf | — | 19% |
Reasoning Kimi K3 leads
GPT-5 Mini: 23.9 (#168), Kimi K3: 63.0 (#17)
| Benchmark | GPT-5 Mini | Kimi K3 |
|---|---|---|
| ARC-AGI-2 | 4.4% | 60.4% |
| ARC-AGI-1 | 54.3% | 94.5% |
| CritPt | 0% | 23.4% |
| Chess Puzzles | 30% | 39% |
| LMArena Hard Prompts | 1380 | 1496 |
| Mystery Game Puzzles | 10% | 26% |
| DTBench | 80.5% | 91.2% |
| LMCA | 34.2% | 52.7% |
| Epoch Capabilities Index | 145.52 | 157.45 |
| ForecastBench | 61 | 61.1 |
| SimpleBench | — | 60.7% |
| Kagi LLM Benchmark | 70.3% | — |
| NYT Connections (extended) | — | 93.6% |
| EnigmaEval | 8.2% | — |
| Surface Evolver Bench | — | 95% |
Math Kimi K3 leads
GPT-5 Mini: 46.7 (#69), Kimi K3: 74.2 (#16)
| Benchmark | GPT-5 Mini | Kimi K3 |
|---|---|---|
| FrontierMath (Tiers 1-3) | 46.7% | 72.2% |
| FrontierMath Tier 4 | 12.2% | 39% |
| OTIS Mock AIME 2024-2025 | 86.7% | 97.2% |
| ProofBench | 9% | 87% |
| LMArena Math | 1378 | 1491 |
| MathArena Final-Answer Competitions | — | 87.8% |
| Omni-MATH | 72.2% | — |
| MATH Level 5 | 97.8% | — |
| FrontierMath (Feb 2025 set) | 27.2% | — |
| FrontierMath Tier 4 (v1) | 6.3% | — |
Knowledge Kimi K3 leads
GPT-5 Mini: 45.6 (#86), Kimi K3: 63.2 (#21)
| Benchmark | GPT-5 Mini | Kimi K3 |
|---|---|---|
| GPQA Diamond | 75% | 93.1% |
| SimpleQA Verified | 21.6% | 50.6% |
| LMArena Expert | 1379 | 1521 |
| Humanity's Last Exam | 19.4% | — |
| MMLU-Pro | 83.5% | — |
| Confabulations | 13.3% | — |
| Vectara Hallucination Rate | 12.9% | — |
| GPQA (HELM) | 75.6% | — |
Multimodal Kimi K3 leads
GPT-5 Mini: 35.6 (#85), Kimi K3: 37.8 (#70)
| Benchmark | GPT-5 Mini | Kimi K3 |
|---|---|---|
| LMArena Vision | 1202 | — |
| VPCT | 40.2% | — |
| Blueprint-Bench 2 | — | 29.5% |
| Furniture Assembly | — | 34.2% |
Multilingual Kimi K3 leads
GPT-5 Mini: 48.9 (#137), Kimi K3: 56.3 (#21)
| Benchmark | GPT-5 Mini | Kimi K3 |
|---|---|---|
| LMArena Non-English | 1363 | 1466 |
| LMArena Chinese | 1385 | 1529 |
| LMArena French | 1386 | 1491 |
| LMArena German | 1366 | 1488 |
| LMArena Japanese | 1341 | 1487 |
| LMArena Korean | 1308 | 1458 |
| LMArena Russian | 1362 | 1482 |
| LMArena Spanish | 1355 | 1472 |
Instruction Following Kimi K3 leads
GPT-5 Mini: 76.2 (#46), Kimi K3: 77.7 (#14)
| Benchmark | GPT-5 Mini | Kimi K3 |
|---|---|---|
| LMArena Instruction Following | 1357 | 1483 |
| IFEval | 92.7% | — |
Long Context Kimi K3 leads
GPT-5 Mini: 41.9 (#132), Kimi K3: 45.8 (#29)
| Benchmark | GPT-5 Mini | Kimi K3 |
|---|---|---|
| LMArena Longer Query | 1355 | 1494 |
| Fiction.LiveBench | 69.4% | — |
Writing & Preference Kimi K3 leads
GPT-5 Mini: 55.2 (#148), Kimi K3: 76.6 (#4)
| Benchmark | GPT-5 Mini | Kimi K3 |
|---|---|---|
| LMArena Text | 1373 | 1476 |
| LMArena Creative Writing | 1325 | 1454 |
| EQ-Bench Creative Writing | 1313 | 2082 |
| LMArena Multi-Turn | 1363 | 1488 |
| Short-Story Creative Writing | 83.1% | — |
| WildBench | 85.5% | — |
| EQ-Bench 4 | — | 1339 |
Frequently asked questions
Is GPT-5 Mini better than Kimi K3?
Kimi K3 is the stronger model overall, scoring 59.5 to 41.8 on the Noometry Index. GPT-5 Mini costs 8.7× 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 Mini or Kimi K3?
GPT-5 Mini is cheaper. It lists at $0.25 per million input tokens and $2 per million output tokens; Kimi K3 lists at $3 and $15.
Is GPT-5 Mini or Kimi K3 better for coding?
Kimi K3 scores higher on coding benchmarks: 61.0 versus 40.1 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 Mini and Kimi K3 share?
37 benchmarks have published results for both models. GPT-5 Mini has 60 scored results on Noometry and Kimi K3 has 53.