Kimi K3vsKimi K2.5
Across 9 shared benchmarks, Kimi K3 leads overall: Kimi K3 wins 9, Kimi K2.5 wins 0, with 0 ties and an average score difference of +94.49.
Kimi K3
Moonshot AI · 2026-07-16 · Reasoning model
Kimi K2.5
Moonshot AI · 2026-01-27 · Multimodal model
Kimi K39 wins(100%)(0%)0 winsKimi K2.5
Benchmark scores
Grouped by capability, sorted by largest gap within each. 9 shared benchmarks.
AI Agent - Information Search
Kimi K3 1/1| Benchmark | Kimi K3 | Kimi K2.5 | Diff |
|---|---|---|---|
| BrowseComp | 91.202 / 56Max (With Tools + Internet) | 60.6037 / 56Thinking (With Tools + Internet) | +30.60 |
AI Agent - Tool Usage
Kimi K3 1/1| Benchmark | Kimi K3 | Kimi K2.5 | Diff |
|---|---|---|---|
| MCP-Atlas | 84.204 / 41Max (With Tools) | 64.4032 / 41Normal (With Tools) | +19.80 |
Coding and Software Engineer
Kimi K3 1/1| Benchmark | Kimi K3 | Kimi K2.5 | Diff |
|---|---|---|---|
| Text Arena (Coding) | 1,6822 / 35Max (No Tools) | 1,43127 / 35Normal (No Tools) | +251.19 |
Commonsense Reasoning
Kimi K3 1/1| Benchmark | Kimi K3 | Kimi K2.5 | Diff |
|---|---|---|---|
| SimpleBench | 60.7029 / 92Max (No Tools) | 46.8052 / 92Thinking (No Tools) | +13.90 |
General Evaluation
Kimi K3 1/1| Benchmark | Kimi K3 | Kimi K2.5 | Diff |
|---|---|---|---|
| GPQA Diamond | 93.5016 / 271Max (No Tools) | 87.6076 / 271Thinking (No Tools) | +5.90 |
General Knowledge
Kimi K3 1/1| Benchmark | Kimi K3 | Kimi K2.5 | Diff |
|---|---|---|---|
| HLE | 5617 / 190Max (With Tools) | 50.2034 / 190Thinking (With Tools) | +5.80 |
Long Context
Kimi K3 1/1| Benchmark | Kimi K3 | Kimi K2.5 | Diff |
|---|---|---|---|
| AA-LCR | 74.708 / 28Max (No Tools) | 6522 / 28Thinking (No Tools) | +9.70 |
Text Embedding
Kimi K3 1/1| Benchmark | Kimi K3 | Kimi K2.5 | Diff |
|---|---|---|---|
| Context Arena | 71.7559 / 126Max (No Tools) | 53.3383 / 126Normal (No Tools) | +18.42 |
Writing and Creative Capabilities
Kimi K3 1/1| Benchmark | Kimi K3 | Kimi K2.5 | Diff |
|---|---|---|---|
| Creative Writing | 2,0712 / 99Normal (No Tools) | 1,57639 / 99Normal (No Tools) | +495.10 |
Specs
| Field | Kimi K3 | Kimi K2.5 |
|---|---|---|
| Publisher | Moonshot AI | Moonshot AI |
| Release date | 2026-07-16 | 2026-01-27 |
| Model type | Reasoning model | Multimodal model |
| Architecture | MoE | MoE |
| Parameters | 2.8T | 1T |
| Context length | 1M | 256K |
| Max output | 1M | 16K |
API pricing
Prices use DataLearner records when available; missing fields are not inferred.
| Item | Kimi K3 | Kimi K2.5 |
|---|---|---|
| Text input | ¥20 / 1M tokens | $0.6 / 1M tokens |
| Text output | ¥100 / 1M tokens | $3 / 1M tokens |
| Cache read | ¥2 / 1M tokens | $0.1 / 1M tokens |
Summary
- Kimi K3leads in:AI Agent - Information Search (1/1), AI Agent - Tool Usage (1/1), Coding and Software Engineer (1/1), Commonsense Reasoning (1/1), General Evaluation (1/1), General Knowledge (1/1), Long Context (1/1), Text Embedding (1/1), Writing and Creative Capabilities (1/1)
On average across the 9 shared benchmarks, Kimi K3 scores 94.49 higher.
Largest single-benchmark gap: Creative Writing — Kimi K3 2,071 vs Kimi K2.5 1,576 (+495.10).
Page generated from structured model, pricing and benchmark records. No real-time LLM is used to write the prose.