Anthropic · released Sep 22, 2026
Claude Opus 5.5
Claude Opus 5.5 is the best all-round model and wins five jobs, from coding to reading images.
Our pick for writingOur pick for codingOur pick for websites & appsOur pick for images & screenshotsOur pick for agents5 weeks oldThinks before answering
- Input price
- $4 / M
- Output price
- $20 / M
- Context
- 1M tokens
- Longest reply
- 128K
- Reads
- text, image, pdf
- Knows up to
- 2026-06
Where it shines
- It tops the web dev arena at 1809, 35 points clear of GPT-6 Astra.
- It ranks first on the agent leaderboard.
- It leads the coding arena at 1545 and vision at 1321.
Where it struggles
- Its search arena score of 1213 trails most rivals.
Where you can use it
- Claude appAppPro, Max, Team and Enterprise plans only. Anthropic raised the 5-hour limits.
- Claude APIAPI$4 in / $20 out per million tokens.
- OpenRouterAPIListed as
anthropic/claude-opus-5.5
How it scores on each job
Scores run 0 to 100 and blend every measure we have for that job. The number after the dot is its rank among the 33 models in the lineup.
The evidence
| Measure | Score | Rank in lineup | Source |
|---|---|---|---|
| Epoch Capabilities Index | 169 | #1 | Epoch AI |
| LMArena text | 1516 | #2 | LMArena |
| LMArena hard prompts | 1536 | #2 | LMArena |
| LMArena instruction following | 1539 | #1 | LMArena |
| LMArena multi-turn | 1504 | #3 | LMArena |
| LMArena creative writing | 1534 | #1 | LMArena |
| LMArena writing & language | 1532 | #1 | LMArena |
| LMArena coding | 1545 | #1 | LMArena |
| LMArena math | 1518 | #4 | LMArena |
| LMArena business & finance | 1497 | #2 | LMArena |
| LMArena legal & government | 1516 | #2 | LMArena |
| LMArena medicine | 1512 | #3 | LMArena |
| WebDev Arena | 1809 | #1 | LMArena |
| Vision Arena | 1321 | #1 | LMArena |
| Search Arena | 1213 | #6 | LMArena |
| Agent Arena | #1 | #1 | LMArena |
| GPQA Diamond | 91% | #18 | Epoch AI |
| FrontierMath | 91% | #3 | Epoch AI |
| SimpleQA Verified | 72% | #5 | Epoch AI |
| ARC-AGI-2 | 93% | #3 | Epoch AI |
Compare it with these
8 weeks old
Anthropic
Claude Fable 5.1
Claude Fable 5.1 reads long documents better than any other model and stays strong at most jobs.
Long documents
- Price
- $10 in · $50 out
- Context
- 1M
4 weeks old
Anthropic
Claude Sonnet 5.5
Claude Sonnet 5.5 does most of what Opus does for half the price, especially agent work.
AgentsWebsites & appsCoding
- Price
- $2 in · $10 out
- Context
- 1M
8 weeks old
OpenAI
GPT-6 Astra
GPT-6 Astra is OpenAI's strongest thinker and the pick for hard math and science problems.
Hard problems
- Price
- $10 in · $50 out
- Context
- 1.1M