Overview
Claude Opus 4.1 is an Opus-tier large language model that Anthropic released on August 5, 2025, as an incremental upgrade to Claude Opus 4. It targets agentic tasks, real-world coding, and reasoning, and on launch it set a new high for the family on SWE-bench Verified at 74.5%, up from 72.5% for Claude Opus 4. Anthropic shipped it at the same price as its predecessor and made it available on the Claude API as claude-opus-4-1-20250805 (alias claude-opus-4-1), as well as on Amazon Bedrock and Google Cloud Vertex AI.
Functionally, Claude Opus 4.1 carries a 200K-token context window, generates up to 32K output tokens, and accepts both text and images (vision). It supports extended thinking, where the model can reason for longer on a request using a thinking budget of up to 64K tokens. Its knowledge is most reliable through January 2025, with training data extending to March 2025. The model focuses its gains on multi-file code refactoring and detail tracking for in-depth research and data analysis rather than on a broad capability jump.
Claude Opus 4.1 is now a legacy model. Anthropic deprecated it and scheduled retirement for August 5, 2026, and recommends migrating to a current Opus model such as Claude Opus 4.8. New projects should generally not start on Opus 4.1; this page documents its verified specifications for teams still running it or evaluating which version they depend on.
| Released | 2025-08-05 |
|---|---|
| License | Proprietary |
| Weights | API only |
| Context | 200K |
| Max output | 32K |
| Architecture | Proprietary transformer-based large language model with optional extended thinking (a configurable reasoning budget of up to 64K tokens). Anthropic has not published parameter counts or architectural details. |
| Knowledge cutoff | Reliable knowledge cutoff January 2025 (training data through March 2025) |
| Modalities | Text, Vision |
| Status | Deprecated (retires August 5, 2026) |
Benchmarks
- SWE-bench Verified74.5%
Scores on a 0–100 scale (25-point gridlines); higher is better. Each benchmark links to its published source.
Pricing
| Input | $15 / 1M tokens per 1M tokens |
|---|---|
| Cached input | $1.50 / 1M tokens (cache read); cache writes $18.75 (5m) / $30 (1h) per 1M per 1M tokens |
| Output | $75 / 1M tokens per 1M tokens |
Batch API is 50% off: $7.50 input / $37.50 output per 1M tokens. Pricing is unchanged from Claude Opus 4.
Strengths
- State-of-the-art real-world coding for its generation at launch — 74.5% on SWE-bench Verified
- Strong agentic and multi-step task execution, including multi-file code refactoring
- Improved detail tracking and agentic search for in-depth research and data analysis
- Optional extended thinking with a budget of up to 64K tokens for harder reasoning problems
- Multimodal: handles both text and image (vision) input
- Available across the Claude API, Amazon Bedrock, and Google Cloud Vertex AI
Best for
- Agentic coding and autonomous multi-step software engineering tasks
- Large multi-file code refactoring and bug fixing
- In-depth research and data analysis that requires careful detail tracking
- Document and image understanding via vision input
- Reasoning-heavy tasks that benefit from extended thinking
How to access
| Provider | Model ID |
|---|---|
| Anthropic (Claude API) ↗ | claude-opus-4-1-20250805 |
| Amazon Bedrock ↗ | anthropic.claude-opus-4-1-20250805-v1:0 |
| Google Cloud Vertex AI ↗ | claude-opus-4-1@20250805 |
Claude Opus — every version
The full lineage of the Claude Opus line, newest first. Every version has its own page — click any to compare specs, benchmarks and pricing.
| Version | Released | Context | License |
|---|---|---|---|
| Claude Opus 4.8current | 2026-05-28 | 1M | Proprietary |
| Claude Opus 4.7 | 2026-04-16 | — | Proprietary |
| Claude Opus 4.6 | 2026-02-05 | — | Proprietary |
| Claude Opus 4.5 | 2025-11-24 | — | Proprietary |
| Claude Opus 4.1 | 2025-08-05 | — | Proprietary |
| Claude Opus 4 | 2025-05-22 | — | Proprietary |
| Claude 3 Opus | 2024-03-04 | — | Proprietary |
FAQ
What is Claude Opus 4.1?
Claude Opus 4.1 is an Opus-tier large language model from Anthropic, released on August 5, 2025, as an incremental upgrade to Claude Opus 4 focused on agentic tasks, real-world coding, and reasoning. Its API model ID is claude-opus-4-1-20250805.
How good is Claude Opus 4.1 at coding?
At launch it set a new state of the art for its generation on SWE-bench Verified with a score of 74.5%, compared with 72.5% for Claude Opus 4. Anthropic also highlighted gains in multi-file code refactoring and agentic coding tasks.
What are the context window and pricing for Claude Opus 4.1?
Claude Opus 4.1 has a 200K-token context window and produces up to 32K output tokens. It is priced at $15 per million input tokens and $75 per million output tokens, the same as Claude Opus 4. The Batch API offers a 50% discount.
Is Claude Opus 4.1 still available?
Claude Opus 4.1 is deprecated and scheduled to retire on August 5, 2026. Anthropic recommends migrating to a current Opus model, such as Claude Opus 4.8, before the retirement date.