Model comparison
Gemini 2.5 Flash vs Kimi K3
Kimi K3 is the stronger model overall, scoring 59.5 to 39.3 on the Noometry Index. Gemini 2.5 Flash costs 7.1× less per token, which makes it the better buy when Kimi K3's lead doesn't matter for your workload.
Last verified . 30 shared benchmarks.
GoogleSummary
- They share 30 benchmarks with published results for both. Gemini 2.5 Flash scores higher in 2 categories and Kimi K3 in 8 categories; 10 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Kimi K3 leads 63.0 to 18.1.
- The biggest single-benchmark swing is ARC-AGI-1: 33.3% for Gemini 2.5 Flash and 94.5% for Kimi K3.
- Gemini 2.5 Flash is cheaper at $0.30 / $2.50 per million input/output tokens, against $3 / $15 for Kimi K3.
- Kimi K3 has downloadable open weights; the other is API-only.
Side by side
| Gemini 2.5 Flash | Kimi K3 | |
|---|---|---|
| Provider | Moonshot AI | |
| Noometry Index | 39.3 | 59.5 |
| Released | 2025-04-17 | 2026-07-16 |
| Weights | Proprietary | Open |
| Context window | 1.05M | 1.05M |
| Max output | 66K | 1.05M |
| Input $ / M tokens | $0.30 | $3 |
| Output $ / M tokens | $2.50 | $15 |
| Results tracked | 54 | 53 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Kimi K3 leads
Gemini 2.5 Flash: 35.8 (#220), Kimi K3: 61.0 (#10)
| Benchmark | Gemini 2.5 Flash | Kimi K3 |
|---|---|---|
| WeirdML | 41.9% | 82.6% |
| LMArena Coding | 1424 | 1508 |
| ALE-Bench | 661.88 | 1,524 |
| DeepSWE | — | 68.5% |
| FrontierCode | — | 44.2% |
| SWE-bench Verified (bash only) | 28.7% | — |
| Aider Polyglot | 55.1% | — |
| LMArena WebDev | — | 1654 |
| FrontierSWE | — | 25.9% |
| SciCode | — | 59.5% |
Agentic & Tool Use Kimi K3 leads
Gemini 2.5 Flash: 30.8 (#74), Kimi K3: 41.8 (#20)
| Benchmark | Gemini 2.5 Flash | Kimi K3 |
|---|---|---|
| Vending-Bench 2 | 548.84 | 5,165 |
| Terminal-Bench | 17.1% | — |
| APEX-Agents | — | 50.6% |
| Berkeley Function Calling Leaderboard | 56.2% | — |
| TheAgentCompany | 41.1% | — |
| τ²-bench Banking | — | 37.1% |
| PostTrainBench | — | 32% |
| BALROG | 33.5% | — |
| GBAEval | — | 48.3% |
| GDP.pdf | — | 19% |
Reasoning Kimi K3 leads
Gemini 2.5 Flash: 18.1 (#286), Kimi K3: 63.0 (#17)
| Benchmark | Gemini 2.5 Flash | Kimi K3 |
|---|---|---|
| ARC-AGI-2 | 2.5% | 60.4% |
| SimpleBench | 41.2% | 60.7% |
| ARC-AGI-1 | 33.3% | 94.5% |
| CritPt | 1.1% | 23.4% |
| LMArena Hard Prompts | 1422 | 1496 |
| DTBench | 76.5% | 91.2% |
| LMCA | 27.5% | 52.7% |
| Epoch Capabilities Index | 143.03 | 157.45 |
| ForecastBench | 60.6 | 61.1 |
| Kagi LLM Benchmark | 56.8% | — |
| NYT Connections (extended) | — | 93.6% |
| Chess Puzzles | — | 39% |
| EnigmaEval | 2.7% | — |
| Mystery Game Puzzles | — | 26% |
| Surface Evolver Bench | — | 95% |
Math Kimi K3 leads
Gemini 2.5 Flash: 39.9 (#98), Kimi K3: 74.2 (#16)
| Benchmark | Gemini 2.5 Flash | Kimi K3 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 73.1% | 97.2% |
| LMArena Math | 1415 | 1491 |
| FrontierMath (Tiers 1-3) | — | 72.2% |
| FrontierMath Tier 4 | — | 39% |
| MathArena Final-Answer Competitions | — | 87.8% |
| ProofBench | — | 87% |
| Omni-MATH | 38.5% | — |
| FrontierMath (Feb 2025 set) | 4.8% | — |
| FrontierMath Tier 4 (v1) | 4.2% | — |
Knowledge Kimi K3 leads
Gemini 2.5 Flash: 36.4 (#168), Kimi K3: 63.2 (#21)
| Benchmark | Gemini 2.5 Flash | Kimi K3 |
|---|---|---|
| LMArena Expert | 1426 | 1521 |
| GPQA Diamond | — | 93.1% |
| Humanity's Last Exam | 12.1% | — |
| SimpleQA Verified | — | 50.6% |
| MMLU-Pro | 63.9% | — |
| Confabulations | 16.8% | — |
| Vectara Hallucination Rate | 7.8% | — |
| GPQA (HELM) | 39% | — |
Multimodal Gemini 2.5 Flash leads
Gemini 2.5 Flash: 41.8 (#32), Kimi K3: 37.8 (#70)
| Benchmark | Gemini 2.5 Flash | Kimi K3 |
|---|---|---|
| LMArena Vision | 1253 | — |
| GeoBench | 76% | — |
| VPCT | 46.2% | — |
| Blueprint-Bench 2 | — | 29.5% |
| Furniture Assembly | — | 34.2% |
| SpatialViz-Bench | 36.9% | — |
Multilingual Kimi K3 leads
Gemini 2.5 Flash: 52.3 (#88), Kimi K3: 56.3 (#21)
| Benchmark | Gemini 2.5 Flash | Kimi K3 |
|---|---|---|
| LMArena Non-English | 1409 | 1466 |
| LMArena Chinese | 1450 | 1529 |
| LMArena French | 1433 | 1491 |
| LMArena German | 1418 | 1488 |
| LMArena Japanese | 1405 | 1487 |
| LMArena Korean | 1385 | 1458 |
| LMArena Russian | 1415 | 1482 |
| LMArena Spanish | 1421 | 1472 |
Instruction Following Kimi K3 leads
Gemini 2.5 Flash: 75.7 (#54), Kimi K3: 77.7 (#14)
| Benchmark | Gemini 2.5 Flash | Kimi K3 |
|---|---|---|
| LMArena Instruction Following | 1405 | 1483 |
| IFEval | 89.8% | — |
Long Context Gemini 2.5 Flash leads
Gemini 2.5 Flash: 47.5 (#17), Kimi K3: 45.8 (#29)
| Benchmark | Gemini 2.5 Flash | Kimi K3 |
|---|---|---|
| LMArena Longer Query | 1419 | 1494 |
| Fiction.LiveBench | 77.8% | — |
Writing & Preference Kimi K3 leads
Gemini 2.5 Flash: 53.8 (#157), Kimi K3: 76.6 (#4)
| Benchmark | Gemini 2.5 Flash | Kimi K3 |
|---|---|---|
| LMArena Text | 1417 | 1476 |
| LMArena Creative Writing | 1400 | 1454 |
| EQ-Bench Creative Writing | 1137 | 2082 |
| LMArena Multi-Turn | 1408 | 1488 |
| Short-Story Creative Writing | 76.5% | — |
| WildBench | 81.7% | — |
| EQ-Bench 4 | — | 1339 |
Frequently asked questions
Is Gemini 2.5 Flash better than Kimi K3?
Kimi K3 is the stronger model overall, scoring 59.5 to 39.3 on the Noometry Index. Gemini 2.5 Flash costs 7.1× less per token, which makes it the better buy when Kimi K3's lead doesn't matter for your workload.
Which is cheaper, Gemini 2.5 Flash or Kimi K3?
Gemini 2.5 Flash is cheaper. It lists at $0.30 per million input tokens and $2.50 per million output tokens; Kimi K3 lists at $3 and $15.
Is Gemini 2.5 Flash or Kimi K3 better for coding?
Kimi K3 scores higher on coding benchmarks: 61.0 versus 35.8 in the Noometry coding category.
Which has the bigger context window?
Both accept 1.05M tokens.
How many benchmarks do Gemini 2.5 Flash and Kimi K3 share?
30 benchmarks have published results for both models. Gemini 2.5 Flash has 54 scored results on Noometry and Kimi K3 has 53.