Kimi K2.6open
Moonshot AI
67.7#51 of 427 overall
- Input / output
- US$ 0,65 / US$ 3,41 US$ per 1M tokens
- Context
- 262K tokens
- Release
- 04/20/2026 release date
- Speed
- 0 tok/s first token in 0.00 s
Family history
The 4 versions of this family, from first to newest, by release date. Click a point to open that version.
See the data as a table
| Model | Released | Score | Rank | AA Intel. | Elo Arena |
|---|---|---|---|---|---|
| Kimi K2 Thinking | 11/06/2025 | 50.2 | #148 | — | 1380 |
| Kimi K2.5 | 01/27/2026 | 64.7 | #64 | — | 1445 |
| Kimi K2.6 | 04/20/2026 | 67.7 | #51 | — | 1456 |
| Kimi K3 | 07/16/2026 | 56.4 | #98 | 17.4 | 1476 |
Scores by source
Scores by sourcepercentile = % of ranked models it beats| Metric | Value | Percentile | Source |
|---|---|---|---|
| Arena text (Elo) weight 2 | 1456 | 81% | LMArena |
| Arena WebDev (Elo) weight 1 | 1509 | 56% | LMArena |
| Arena vision (Elo) weight 0.5 | 1282 | 75% | LMArena |
| Humanity's Last Exam weight 1 | 37.5 | 86% | Artificial Analysis |
| Terminal-Bench weight 1 | 0.5 | 49% | Artificial Analysis |
| GPQA Diamond | 91.1 | 93% | Artificial Analysis |
| Long-context reasoning | 81.0 | 90% | Artificial Analysis |
By category
| Category | Score | Rank | Metrics |
|---|---|---|---|
| Agentic | 49.3 | #65 | Terminal-Bench |
| Coding | 53.0 | #48 | Arena WebDev (Elo) |
| Reasoning | 76.2 | #36 | Humanity's Last Exam, GPQA Diamond |
| Research | 70.0 | #31 | Long-context reasoning |
| Design | 53.0 | #47 | Arena WebDev (Elo) |
| Chat | 65.5 | #38 | Arena text (Elo) |
| Vision | 62.6 | #27 | Arena vision (Elo) |
History
RankingIA score on each collection day (hover to see the rank).
Daily collection started on 10/30/2026. The chart appears from the second day the model is ranked.