Compare / head-to-head
Gemini 3 Flash Previewvs
Kimi K2.5
Gemini 3 Flash Preview leads 11 of 12 shared benchmarks. Gemini 3 Flash Preview has the larger context window (1.0M tokens).
Benchmarks from cited public sources; pricing from official pages; status from official provider feeds.
Shared benchmarks
11 – 1
Gemini 3 Flash Preview leads
Cheaper per token
—
list price, input + output
Larger context
Gemini 3 Flash Preview
1.0M tokens
Providers
2 providers
Google · Moonshot AI
Gemini 3 Flash Preview
Google · released 2025-12-17
textvisionaudio
- Context
- 1.0M
- Max out
- 66K
- Input /1M
- —
- Output /1M
- —
- Cached /1M
- —
- Scores
- 34 · 12 core
Kimi K2.5
Moonshot AI · released 2026-01-27
textvisionvideo
- Context
- 256K
- Max out
- —
- Input /1M
- —
- Output /1M
- —
- Cached /1M
- —
- Scores
- 26 · 14 core
Quality
Benchmark matrix
12 shared · 14 only Gemini 3 Flash Preview · 14 only Kimi K2.5| Benchmark | Gemini 3 Flash Preview | Kimi K2.5 | Δ | Edge |
|---|---|---|---|---|
| Reported by both · 12 | ||||
| SWE-bench Verified | 78% ↗ | 76.8% ↗ | +1.2 pt | Gemini 3 Flash Preview |
| AIME 2025 | 95.2% ↗ | 96.1% ↗ | -0.9 pt | Kimi K2.5 |
| LMArena Elo | 1472.7 ↗ | 1450.1 ↗ | +22.6 | Gemini 3 Flash Preview |
| LMArena Vision (LMArena) | 1272.9 ↗ | 1252.2 ↗ | +20.7 | Gemini 3 Flash Preview |
| LMArena WebDev (LMArena) | 1438.9 ↗ | 1435.8 ↗ | +3.1 | Gemini 3 Flash Preview |
| MMMU-Pro | 81.2% ↗ | 78.5% ↗ | +2.7 pt | Gemini 3 Flash Preview |
| SAGE (Vals AI) | 51.8% ↗ | 49.9% ↗ | +1.9 pt | Gemini 3 Flash Preview |
| SimpleBench (SimpleBench) | 61.1% ↗ | 46.8% ↗ | +14.3 pt | Gemini 3 Flash Preview |
| SimpleQA Verified | 68.7% ↗ | 34.3% ↗epoch run | +34.4 pt | Gemini 3 Flash Preview |
| SWE-bench Verified (Epoch AI run) | 75.4% ↗epoch run | 73.8% ↗epoch run | +1.6 pt | Gemini 3 Flash Preview |
| Vending-Bench 2 (Andon Labs) | 3634.72 ↗ | 1198.46 ↗ | +2436.3 | Gemini 3 Flash Preview |
| WeirdML (Håvard Tveit Ihle) | 61.6% ↗ | 45.6% ↗ | +16 pt | Gemini 3 Flash Preview |
| Only Gemini 3 Flash Preview reports · 14 | ||||
| GPQA Diamond | 90.4% ↗ | not reported | — | — |
| Humanity's Last Exam (no tools) | 33.7% ↗ | not reported | — | — |
| BALROG (BALROG) | 48.1% ↗ | not reported | — | — |
| Chess Puzzles (Epoch AI run) | 40% ↗epoch run | not reported | — | — |
| FrontierMath Tier 4 v2 (Epoch AI run) | 17.1% ↗epoch run | not reported | — | — |
| FrontierMath Tiers 1-3 v2 (Epoch AI run) | 51.2% ↗epoch run | not reported | — | — |
| GSO Opt@1 (GSO) | 7.8% ↗ | not reported | — | — |
| Mystery Game Puzzles (Epoch AI run) | 26% ↗epoch run | not reported | — | — |
| OTIS Mock AIME 2024-2025 (Epoch AI run) | 95.6% ↗epoch run | not reported | — | — |
| tau2-bench Airline (Sierra) | 82.5% ↗ | not reported | — | — |
| tau2-bench Banking Knowledge (Sierra) | 27.3% ↗ | not reported | — | — |
| tau2-bench Retail (Sierra) | 76.8% ↗ | not reported | — | — |
| tau2-bench Telecom (Sierra) | 91.2% ↗ | not reported | — | — |
| Terminal-Bench 2.0 | 47.6% ↗ | not reported | — | — |
| Only Kimi K2.5 reports · 14 | ||||
| MMLU-Pro | not reported | 87.1% ↗ | — | — |
| BrowseComp | not reported | 60.6% ↗ | — | — |
| GPQA-Diamond | not reported | 87.6% ↗ | — | — |
| HLE-Full | not reported | 30.1% ↗ | — | — |
| HLE-Full (w/ tools) | not reported | 50.2% ↗ | — | — |
| HMMT 2025 (Feb) | not reported | 95.4% ↗ | — | — |
| IMO-AnswerBench | not reported | 81.8% ↗ | — | — |
| LiveCodeBench (v6) | not reported | 85% ↗ | — | — |
| OSWorld-Verified (XLANG) | not reported | 63.3% ↗ | — | — |
| SWE-Bench Multilingual | not reported | 73% ↗ | — | — |
| SWE-Bench Pro | not reported | 50.7% ↗ | — | — |
| Terminal Bench 2.0 | not reported | 50.8% ↗ | — | — |
| Toolathlon-Verified (HKUST) | not reported | 33% ↗ | — | — |
| VideoMMMU | not reported | 86.6% ↗ | — | — |
Scores tagged "epoch" or "matharena" are independent runs, used only where the lab has not published its own; ⚠ marks rows MathArena flags as released after the competition. Higher is better on every row. Δ is Gemini 3 Flash Preview minus Kimi K2.5 in the benchmark's own unit. "Not reported" means the lab has not published that figure; it is not a zero. ↗ opens the source.
Specs & pricing
Side by side
Official pricing pages and model cards| Spec | Gemini 3 Flash Preview | Kimi K2.5 | Edge |
|---|---|---|---|
| Context window | 1.0M tokens | 256K tokens | Gemini 3 Flash Preview |
| Max output | 66K tokens | — | — |
| Input price / 1M | — | — | — |
| Output price / 1M | — | — | — |
| Cached input / 1M | — | — | — |
| Input + output / 1M Lower is cheaper. List prices; batch, tool and regional fees excluded. | — | — | — |
| Modalities | text · vision · audio | text · vision · video | Tie |
| Released | 2025-12-17 | 2026-01-27 | — |
| Cited benchmark scores | 34 | 26 | — |
Reliability
Provider status
Live from /statusMore matchups
Gemini 3 Flash Preview vs …
Models sharing the most benchmarksMore matchups
Kimi K2.5 vs …
Models sharing the most benchmarksBuilt by Respan
Which one wins on your data?
Public benchmarks are a starting point. Run Gemini 3 Flash Preview and Kimi K2.5 on your own prompts with Respan evals, or route to either through one gateway key with automatic failover.