Summary
On July 24, 2026, Anthropic released Claude Opus 5, a frontier model priced at $5/$25 per million input/output tokens (unchanged from Opus 4.8 and roughly half the cost of Fable 5), with a 1M-token context window, 128K max output, and a May 2026 knowledge cutoff. Within days it reached major enterprise platforms: available as a Databricks-hosted model on July 24, added to Amazon Bedrock by the July 27 AWS roundup, and offered in GitHub Copilot for agentic coding, alongside its rollout as the default on Claude Max and top model on Claude Pro.
What changed
Anthropic shipped Claude Opus 5 (1M-token context, 128K max output, $5/$25 pricing, Auto Mode safety scanning) and it rapidly became available across enterprise surfaces: Databricks-hosted model serving (July 24), Amazon Bedrock (by July 27), and GitHub Copilot, in addition to Claude Max and Claude Pro.
Why it matters
Holding Opus-tier pricing flat while nearly matching Fable 5 quality resets the price-performance frontier for agentic coding and long-context work. The fast fan-out to Databricks, AWS Bedrock, and GitHub Copilot matters as much as the model itself: enterprises can adopt Opus 5 inside their existing data and developer platforms without leaving governed environments, accelerating displacement of pricier incumbents in production agent and coding workloads.
Evidence excerpt
Claude Opus 5 launched July 24, 2026 at $5/$25 per million input/output tokens (same as Opus 4.8, half of Fable 5's $10/$50), with a 1M-token context window ... 'Databricks Model Serving now supports Anthropic Claude Opus 5 as a Databricks-hosted model' (July 24, 2026) ... AWS Weekly Roundup (July 27, 2026): Claude Opus 5 on AWS.