Anthropic · released Sep 28, 2026
Claude Sonnet 5.5
Claude Sonnet 5.5 does most of what Opus does for half the price, especially agent work.
4 weeks oldThinks before answering
- Input price
- $2 / M
- Output price
- $10 / M
- Context
- 1M tokens
- Longest reply
- 128K
- Reads
- text, image, pdf
- Knows up to
- 2026-06
Where it shines
- It ranks second on the agent leaderboard.
- It scores 1769 in web dev, close to GPT-6 Astra.
- It costs $2 in, $10 out per million tokens.
Where it struggles
- Its SimpleQA score of 46.5 means it gets more facts wrong.
Where you can use it
- Anthropic APIAPI$2 in / $10 out per million tokens.
- OpenRouterAPIListed as
anthropic/claude-sonnet-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 | 166 | #6 | Epoch AI |
| LMArena text | 1473 | #13 | LMArena |
| LMArena hard prompts | 1493 | #11 | LMArena |
| LMArena instruction following | 1491 | #6 | LMArena |
| LMArena multi-turn | 1465 | #16 | LMArena |
| LMArena creative writing | 1468 | #9 | LMArena |
| LMArena writing & language | 1476 | #9 | LMArena |
| LMArena coding | 1510 | #7 | LMArena |
| LMArena math | 1498 | #10 | LMArena |
| LMArena business & finance | 1468 | #13 | LMArena |
| LMArena legal & government | 1480 | #14 | LMArena |
| LMArena medicine | 1467 | #16 | LMArena |
| WebDev Arena | 1769 | #3 | LMArena |
| Vision Arena | 1287 | #12 | LMArena |
| Agent Arena | #2 | #2 | LMArena |
| GPQA Diamond | 96% | #3 | Epoch AI |
| Humanity's Last Exam | 46% | #4 | Epoch AI |
| FrontierMath | 89% | #6 | Epoch AI |
| SimpleQA Verified | 47% | #15 | Epoch AI |
| ARC-AGI-2 | 84% | #9 | 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
New
Gemini 4 Argon
Gemini 4 Argon is Google's new flagship and gives the best answers for everyday chat and research.
Everyday questionsResearchProfessional work
- Price
- $2 in · $10 out
- Context
- 1M
5 weeks old
Anthropic
Claude Opus 5.5
Claude Opus 5.5 is the best all-round model and wins five jobs, from coding to reading images.
WritingCodingWebsites & apps
- Price
- $4 in · $20 out
- Context
- 1M