Claude Opus 5 is an Anthropic model released on July 24, 2026 for complex agentic coding and professional work. Anthropic describes it as a step-change improvement over Opus 4.8, with a long context window and adaptive thinking enabled by default.
What is Claude Opus 5?
Opus 5 belongs to the Opus tier of the Claude model family. It is designed for long-running coding agents, large-scale refactoring, advanced research, knowledge work, vision-heavy workflows, and tasks that combine tool use with text and image understanding.
Context window and output
Claude Opus 5 supports a one-million-token context window as both its default and maximum context size. It can generate up to 128,000 output tokens. Adaptive thinking is enabled by default, while the effort setting controls how much reasoning the model uses.
Modalities and tools
The model accepts text and images and produces text. Anthropic documents its use in tool-driven workflows, multistep coding, document work, and computer use. These capabilities do not guarantee correctness; important outputs and tool actions still require review.
API pricing
Standard pricing is $5 per million input tokens and $25 per million output tokens. Cache hits cost $0.50 per million tokens. A five-minute cache write costs $6.25, while a one-hour cache write costs $10 per million tokens.
Strengths
- A one-million-token context window.
- Strong long-horizon agentic coding and professional-work capabilities.
- Adaptive thinking with image and tool support.
Limitations
- Anthropic positions Claude Fable 5 as its more capable widely available model.
- Anthropic reports that Opus 5 trails Mythos 5 on cybersecurity tasks.
- The newer tokenizer can produce more tokens for the same text, affecting usage calculations.
In short
Claude Opus 5 is designed for demanding coding, reasoning, and long-context work. It offers lower pricing than Anthropic’s highest-capability tier, but it is neither the cheapest nor the best choice for every workload.