Overview
Claude Opus 5.5 is Anthropic’s model for long-running agentic coding and knowledge work. It was released on September 22, 2026. Anthropic reports improvements in agentic coding, computer use, knowledge work, and communication compared with earlier models. (anthropic.com)
Capabilities and uses
- Accepts text and image input and produces text output.
- Supports a 1 million-token context window and up to 128,000 output tokens. The Message Batches API supports up to 300,000 output tokens with a specified beta header.
- Uses adaptive thinking, with
mediumas the default effort level. - Intended uses include multi-hour coding agents, large-scale refactoring, complex systems engineering, visual workflows, computer use, and professional knowledge work.
- Available through the Claude API and other listed cloud platforms, including Amazon Bedrock, Google Cloud, and Microsoft Foundry. (platform.claude.com)
Limitations
The reliable knowledge and training-data cutoff is June 2026. Adaptive thinking is always enabled, although its depth can be controlled using the effort parameter. Outputs can be incorrect and should be checked, particularly before consequential or automated actions. Anthropic notes that benchmark differences are not always reliable indicators of real-world differences and that safety safeguards can intervene in high-risk areas. (platform.claude.com)
Claude API pricing
The following are Anthropic’s USD list prices per 1 million tokens, covering all context-window sizes. The official rate card is effective November 9, 2026; checked November 16, 2026. Other platforms and tools may have different prices. (assets.claude.com)
| Tier and inference scope | Input | Output | 5-minute cache write | 1-hour cache write | Cache hit/refresh |
|---|---|---|---|---|---|
| Standard, global | $4.00 | $20.00 | $5.00 | $8.00 | $0.20 |
| Standard, US-only inference | $4.40 | $22.00 | $5.50 | $8.80 | $0.22 |
| Batch, global | $2.00 | $10.00 | $2.50 | $4.00 | $0.10 |
| Batch, US-only inference | $2.20 | $11.00 | $2.75 | $4.40 | $0.11 |
| Fast mode, global | $8.00 | $40.00 | $10.00 | $16.00 | $0.40 |
| Fast mode, US-only inference | $8.80 | $44.00 | $11.00 | $17.60 | $0.44 |
Batch and fast-mode rates apply only when those modes are used. US-only inference is a separate scope from global inference. Anthropic notes that promotional prices, when offered, are subject to their own terms. (assets.claude.com)