Kimi K2.5vsKimi K2 0905
Across 4 shared benchmarks, Kimi K2.5 leads overall: Kimi K2.5 wins 4, Kimi K2 0905 wins 0, with 0 ties and an average score difference of +20.01.
Kimi K2.5
Moonshot AI · 2026-01-27 · Multimodal model
Kimi K2 0905
Moonshot AI · 2025-09-05 · Chat model
Kimi K2.54 wins(100%)(0%)0 winsKimi K2 0905
Benchmark scores
Grouped by capability, sorted by largest gap within each. 4 shared benchmarks.
Coding and Software Engineer
Kimi K2.5 2/2| Benchmark | Kimi K2.5 | Kimi K2 0905 | Diff |
|---|---|---|---|
| SWE-Bench Pro - Public | 50.7031 / 43Thinking (With Tools) | 27.6743 / 43 | +23.03 |
| SWE-bench Verified | 76.8027 / 108Thinking (With Tools) | 69.2059 / 108 | +7.60 |
General Knowledge
Kimi K2.5 1/1| Benchmark | Kimi K2.5 | Kimi K2 0905 | Diff |
|---|---|---|---|
| HLE | 50.2020 / 157Thinking (With Tools) | 21.7095 / 157 | +28.50 |
Math and Reasoning
Kimi K2.5 1/1| Benchmark | Kimi K2.5 | Kimi K2 0905 | Diff |
|---|---|---|---|
| AIME2025 | 96.1021 / 106Thinking (No Tools) | 75.2067 / 106 | +20.90 |
Specs
| Field | Kimi K2.5 | Kimi K2 0905 |
|---|---|---|
| Publisher | Moonshot AI | Moonshot AI |
| Release date | 2026-01-27 | 2025-09-05 |
| Model type | Multimodal model | Chat model |
| Architecture | MoE | MoE |
| Parameters | 1T | 1T |
| Context length | 256K | 256K |
| Max output | 16K | 4K |
Summary
- Kimi K2.5leads in:Coding and Software Engineer (2/2), General Knowledge (1/1), Math and Reasoning (1/1)
On average across the 4 shared benchmarks, Kimi K2.5 scores 20.01 higher.
Largest single-benchmark gap: HLE — Kimi K2.5 50.20 vs Kimi K2 0905 21.70 (+28.50).
Page generated from structured model, pricing and benchmark records. No real-time LLM is used to write the prose.