SI Model · Anthropic
Claude Opus 5.5
On this page
| Developer | Anthropic |
|---|---|
| Released | September 22, 2026 |
| Status | Current |
| Inputs | Text, Images |
| Context window | 1M tokens |
| API price (input) | $4 / 1M tokens |
| API price (output) | $20 / 1M tokens |
| Open weights | No |
| Access | Claude apps, Claude API, Amazon Bedrock, Google Cloud, Microsoft Foundry |
Specs and list prices last verified October 7, 2026. Prices change often; check the provider before you buy.
Strengths
- Anthropic's recommended starting point for most workloads
- Long-running agentic coding
- 1M-token context and 128K output
- Cheaper prompt caching (5% of input price)
Best for
- Software engineering and coding agents
- Knowledge work such as analysis and writing
- Most production workloads that need top-tier quality
Claude Opus 5.5 is the model Anthropic tells developers to “start with”. Launched on September 22, 2026, it is built for long-running agentic coding and knowledge work, and sits between the premium Claude Fable 5.1 and the faster Claude Sonnet 5.5.
Key specs
- Context window: 1 million tokens; up to 128,000 output tokens (more via the Batch API)
- Thinking: adaptive and always on, with “medium” default effort
- Price: $4 input / $20 output per million tokens; prompt-cache reads at 5% of input price; 50% off for batch
- Knowledge cutoff: June 2026
Why it matters
Opus 5.5 brings near-flagship capability at well under half the price of Fable 5.1 or GPT-6 Astra. That makes it a common choice for coding agents that consume huge numbers of tokens.
Compare
Sources
- Claude models overview — Anthropic
Written by
Luka Kušec · Editor
Editor of SI.info. Writes about Super Intelligence, technology policy and the people building frontier models.