Model comparison
Gemini 2.5 Flash vs Kimi K2.5
Kimi K2.5 is the stronger model overall, scoring 48.1 to 39.3 on the Noometry Index.
Last verified . 37 shared benchmarks.
GoogleSummary
- They share 37 benchmarks with published results for both. Gemini 2.5 Flash scores higher in 2 categories and Kimi K2.5 in 8 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where Kimi K2.5 leads 53.6 to 36.4.
- The biggest single-benchmark swing is SWE-bench Verified (bash only): 28.7% for Gemini 2.5 Flash and 70.8% for Kimi K2.5.
- Gemini 2.5 Flash is cheaper at $0.30 / $2.50 per million input/output tokens, against $0.45 / $2.25 for Kimi K2.5.
- Gemini 2.5 Flash accepts more context: 1.05M tokens versus 262K.
- Kimi K2.5 has downloadable open weights; the other is API-only.
Side by side
| Gemini 2.5 Flash | Kimi K2.5 | |
|---|---|---|
| Provider | Moonshot AI | |
| Noometry Index | 39.3 | 48.1 |
| Released | 2025-04-17 | 2026-01-27 |
| Weights | Proprietary | Open |
| Context window | 1.05M | 262K |
| Max output | 66K | 262K |
| Input $ / M tokens | $0.30 | $0.45 |
| Output $ / M tokens | $2.50 | $2.25 |
| Results tracked | 54 | 51 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Kimi K2.5 leads
Gemini 2.5 Flash: 35.8 (#220), Kimi K2.5: 48.8 (#53)
| Benchmark | Gemini 2.5 Flash | Kimi K2.5 |
|---|---|---|
| SWE-bench Verified (bash only) | 28.7% | 70.8% |
| WeirdML | 41.9% | 45.6% |
| LMArena Coding | 1424 | 1474 |
| ALE-Bench | 661.88 | 821.65 |
| SWE-bench Verified | — | 73.8% |
| Aider Polyglot | 55.1% | — |
| LMArena WebDev | — | 1437 |
| SWE-bench Multilingual | — | 67.3% |
| SciCode | — | 49% |
Agentic & Tool Use Kimi K2.5 leads
Gemini 2.5 Flash: 30.8 (#74), Kimi K2.5: 34.2 (#48)
| Benchmark | Gemini 2.5 Flash | Kimi K2.5 |
|---|---|---|
| Terminal-Bench | 17.1% | 43.2% |
| Vending-Bench 2 | 548.84 | 1,198 |
| Berkeley Function Calling Leaderboard | 56.2% | — |
| TheAgentCompany | 41.1% | — |
| OSWorld | — | 63.3% |
| BALROG | 33.5% | — |
Reasoning Kimi K2.5 leads
Gemini 2.5 Flash: 18.1 (#286), Kimi K2.5: 31.2 (#80)
| Benchmark | Gemini 2.5 Flash | Kimi K2.5 |
|---|---|---|
| ARC-AGI-2 | 2.5% | 11.8% |
| SimpleBench | 41.2% | 46.8% |
| Kagi LLM Benchmark | 56.8% | 78.5% |
| ARC-AGI-1 | 33.3% | 65.3% |
| CritPt | 1.1% | 3.1% |
| EnigmaEval | 2.7% | 3.4% |
| LMArena Hard Prompts | 1422 | 1453 |
| Epoch Capabilities Index | 143.03 | 148.03 |
| NYT Connections (extended) | — | 69.9% |
| Chess Puzzles | — | 12% |
| Thematic Generalization | — | 69.4% |
| DTBench | 76.5% | — |
| LMCA | 27.5% | — |
| ForecastBench | 60.6 | — |
Math Kimi K2.5 leads
Gemini 2.5 Flash: 39.9 (#98), Kimi K2.5: 51.8 (#53)
| Benchmark | Gemini 2.5 Flash | Kimi K2.5 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 73.1% | 92.2% |
| LMArena Math | 1415 | 1470 |
| FrontierMath (Feb 2025 set) | 4.8% | 27.9% |
| FrontierMath Tier 4 (v1) | 4.2% | 4.2% |
| MathArena Final-Answer Competitions | — | 62.3% |
| Omni-MATH | 38.5% | — |
Knowledge Kimi K2.5 leads
Gemini 2.5 Flash: 36.4 (#168), Kimi K2.5: 53.6 (#56)
| Benchmark | Gemini 2.5 Flash | Kimi K2.5 |
|---|---|---|
| Humanity's Last Exam | 12.1% | 24.4% |
| Vectara Hallucination Rate | 7.8% | 14.2% |
| LMArena Expert | 1426 | 1466 |
| GPQA Diamond | — | 87.6% |
| SimpleQA Verified | — | 34.3% |
| MMLU-Pro | 63.9% | — |
| Confabulations | 16.8% | — |
| GPQA (HELM) | 39% | — |
Multimodal Too close to call
Gemini 2.5 Flash: 41.8 (#32), Kimi K2.5: 41.1 (#39)
| Benchmark | Gemini 2.5 Flash | Kimi K2.5 |
|---|---|---|
| LMArena Vision | 1253 | 1269 |
| GeoBench | 76% | — |
| VPCT | 46.2% | — |
| LMArena Document | — | 1430 |
| SpatialViz-Bench | 36.9% | — |
Multilingual Kimi K2.5 leads
Gemini 2.5 Flash: 52.3 (#88), Kimi K2.5: 53.9 (#53)
| Benchmark | Gemini 2.5 Flash | Kimi K2.5 |
|---|---|---|
| LMArena Non-English | 1409 | 1433 |
| LMArena Chinese | 1450 | 1495 |
| LMArena French | 1433 | 1454 |
| LMArena German | 1418 | 1441 |
| LMArena Japanese | 1405 | 1421 |
| LMArena Korean | 1385 | 1410 |
| LMArena Russian | 1415 | 1435 |
| LMArena Spanish | 1421 | 1450 |
Instruction Following Too close to call
Gemini 2.5 Flash: 75.7 (#54), Kimi K2.5: 75.3 (#64)
| Benchmark | Gemini 2.5 Flash | Kimi K2.5 |
|---|---|---|
| LMArena Instruction Following | 1405 | 1431 |
| IFEval | 89.8% | — |
Long Context Kimi K2.5 leads
Gemini 2.5 Flash: 47.5 (#17), Kimi K2.5: 52.1 (#7)
| Benchmark | Gemini 2.5 Flash | Kimi K2.5 |
|---|---|---|
| Fiction.LiveBench | 77.8% | 86.1% |
| LMArena Longer Query | 1419 | 1445 |
| CL-bench | — | 19.3% |
| CL-bench Life | — | 13.2% |
Writing & Preference Kimi K2.5 leads
Gemini 2.5 Flash: 53.8 (#157), Kimi K2.5: 65.1 (#53)
| Benchmark | Gemini 2.5 Flash | Kimi K2.5 |
|---|---|---|
| LMArena Text | 1417 | 1445 |
| LMArena Creative Writing | 1400 | 1423 |
| EQ-Bench Creative Writing | 1137 | 1579 |
| LMArena Multi-Turn | 1408 | 1444 |
| Short-Story Creative Writing | 76.5% | — |
| WildBench | 81.7% | — |
Frequently asked questions
Is Gemini 2.5 Flash better than Kimi K2.5?
Kimi K2.5 is the stronger model overall, scoring 48.1 to 39.3 on the Noometry Index.
Which is cheaper, Gemini 2.5 Flash or Kimi K2.5?
Gemini 2.5 Flash is cheaper. It lists at $0.30 per million input tokens and $2.50 per million output tokens; Kimi K2.5 lists at $0.45 and $2.25.
Is Gemini 2.5 Flash or Kimi K2.5 better for coding?
Kimi K2.5 scores higher on coding benchmarks: 48.8 versus 35.8 in the Noometry coding category.
Which has the bigger context window?
Gemini 2.5 Flash does, with 1.05M tokens against 262K.
How many benchmarks do Gemini 2.5 Flash and Kimi K2.5 share?
37 benchmarks have published results for both models. Gemini 2.5 Flash has 54 scored results on Noometry and Kimi K2.5 has 51.