Claude Opus 5.5
Anthropic
93.5#1 of 427 overall
- Input / output
- US$ 4,00 / US$ 20,00 US$ per 1M tokens
- Context
- 1M tokens
- Release
- 09/22/2026 release date
- Weights
- Closed
- Speed
- 97 tok/s first token in 562.71 s
Family history
The 9 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 |
|---|---|---|---|---|---|
| Claude 3 Opus | 03/04/2024 | 14.9 | #424 | — | 1262 |
| Claude 4 Opus | 05/22/2025 | 43.8 | #210 | — | 1377 |
| Claude Opus 4.1 | 08/05/2025 | 49.7 | #150 | — | 1419 |
| Claude Opus 4.5 | 11/24/2025 | 68.1 | #49 | — | 1451 |
| Claude Opus 4.6 | 02/04/2026 | 81.3 | #14 | — | 1502 |
| Claude Opus 4.7 | 04/16/2026 | 81.2 | #15 | — | 1490 |
| Claude Opus 4.8 | 05/27/2026 | 77.1 | #20 | — | 1461 |
| Claude Opus 5 | 07/24/2026 | 89.1 | #4 | — | 1508 |
| Claude Opus 5.5 | 09/22/2026 | 93.5 | #1 | 58.4 | 1516 |
Scores by source
Scores by sourcepercentile = % of ranked models it beats| Metric | Value | Percentile | Source |
|---|---|---|---|
| LiveBench weight 3 | 85.1 | 94% | LiveBench |
| AA Intelligence Index weight 3 | 58.4 | 98% | Artificial Analysis |
| Arena text (Elo) weight 2 | 1516 | 99% | LMArena |
| Arena WebDev (Elo) weight 1 | 1810 | 100% | LMArena |
| Arena vision (Elo) weight 0.5 | 1321 | 99% | LMArena |
| Humanity's Last Exam weight 1 | 61.4 | 100% | Artificial Analysis |
| Terminal-Bench weight 1 | 59.6 | 98% | Artificial Analysis |
| Long-context reasoning | 84.7 | 99% | Artificial Analysis |
| LiveBench long agentic | 72.8 | 94% | LiveBench |
| LiveBench reasoning | 92.2 | 90% | LiveBench |
| LiveBench coding | 96.3 | 97% | LiveBench |
| LiveBench browser | 88.4 | 84% | LiveBench |
| LiveBench data analysis | 80.3 | 82% | LiveBench |
| LiveBench research | 85.0 | 90% | LiveBench |
| LiveBench design | 86.9 | 100% | LiveBench |
By category
| Category | Score | Rank | Metrics |
|---|---|---|---|
| Agentic | 80.5 | #3 | LiveBench long agentic, Terminal-Bench |
| Coding | 82.3 | #1 | LiveBench coding, Arena WebDev (Elo) |
| Reasoning | 80.1 | #13 | LiveBench reasoning, Humanity's Last Exam |
| Data analysis | 66.2 | #6 | LiveBench data analysis |
| Research | 79.8 | #1 | LiveBench research, Long-context reasoning |
| Design | 83.3 | #1 | LiveBench design, Arena WebDev (Elo) |
| Browser use | 67.0 | #6 | LiveBench browser |
| Chat | 74.5 | #3 | Arena text (Elo) |
| Vision | 74.5 | #2 | 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.