Anthropic shipped Claude Opus 5 on July 24, its fourth model in two months. It’s a near-frontier LLM: at max effort it comes within 0.5% of flagship Fable 5’s peak on the CursorBench coding eval, but costs $5/$25 per million input/output tokens — roughly half the cost per task. Day one, it topped HackerNews (1,381 points) and Artificial Analysis’s intelligence leaderboard.
What it actually is
A general-purpose model tuned for the boring-but-valuable work: agentic coding, tool-calling agents, knowledge tasks. The new low/medium/high effort dial is the real trick — you pay for reasoning only when the job needs it. Cheap for a summary, expensive for a multi-step agent, your call per request. It ships with a 1M-token context window, so whole codebases fit in one prompt, and it’s now the default model in Claude Max.
API and where to plug in
claude-opus-5 is live on the Claude API, Amazon Bedrock, Google Cloud, and Microsoft Foundry, plus claude.ai, Claude Code, and Cowork. Point your existing Anthropic SDK at the new model ID and the effort toggle becomes a per-request parameter — same code, cheaper bill.
The bet is obvious: most work doesn’t need the flagship. Frontier-adjacent quality at commodity pricing is how you make a model the office default.
You Might Also Like
- 13b Into Openai yet Microsoft Copilot Cowork Runs on Anthropic Claude
- Anthropic put Claude Cowork on Your Phone and 90 of Users Arent Coding
- Claude for Government Claude Code Cowork Enter Public Beta on Fedramp High
- Claude Cowork Cloud Anthropic Close Your Laptop the Agent Keeps Working
- Claude Code Remote Control Just Turned my Phone Into a Coding Terminal and im Weirdly Into it
Top AI Product 
Leave a comment