Enterprise

Anthropic launches Claude Opus 5 as its everyday enterprise model at half of Fable 5's price

Opus 5 becomes the default on Claude Max and ships across every paid plan at $5/$25 per million tokens, matching Opus 4.8 pricing while beating Fable 5 on a headline coding benchmark.

By Asher Glick · Enterprise software desk · 2026-07-26 · 3 min read

Photo: Unsplash / Google DeepMind — Abstract swirling orange and pink AI-generated pattern representing large language models

Anthropic released Claude Opus 5 on Friday at $5 per million input tokens and $25 per million output, matching Opus 4.8’s pricing and undercutting its own Fable 5 flagship by half. The framing is the story: a near-frontier model repositioned as the everyday default rather than the premium tier, priced identically to the model it displaces.

Opus 5 becomes the default on Claude Max and ships across every paid plan, with distribution live on day one through the Claude API, Amazon Bedrock, Google Cloud, Microsoft Foundry, claude.ai, Claude Code, and Cowork. It carries a 1M token context window, thinking on by default, an effort dial, mid-task model switching, and Batch API pricing at half list ($2.50 input, $12.50 output, with cache hits at $0.50 per MTok).

The cadence is what to watch. Opus 4.8 arrived May 28. Mythos 5, Fable 5, and Sonnet 5 landed in June. Two months later, Opus 5 completes the 5-series across every tier except Haiku. On the internal Frontier Bench v0.1 agentic coding evaluation, Anthropic reports Opus 5 at 43.3%, ahead of GPT-5.6 Sol at 34.4%, Fable 5 at 33.7%, and Opus 4.8 at 21.1%. Box, cited in the launch materials, reports Opus 5 outperforming Opus 4.8 by 8% overall, with 11% gains in data analysis and 17% in due diligence workflows.

The enterprise pitch has teeth beyond the benchmark. Opus 5 is exempt from the 30-day data retention policy applied to Fable 5 and Mythos 5, and Anthropic expects its safety classifiers to engage 85% less often than on Fable 5. Cybersecurity guardrails remain: the model is blocked from scanning binaries for vulnerabilities, though it can read source code.

The competitive backdrop is louder than the release itself. Moonshot’s open-weight Kimi K3 has been pressuring commercial pricing from below, and the Trump administration has moved to delay some model releases pending review. An Anthropic spokesperson told Axios: “We continue to work with our government partners to conduct their own independent testing of our models. This includes Opus 5.”

Pricing the second-best model at the price of last quarter’s best is a familiar move from cloud infrastructure’s commodity phase. Anthropic is running the same play at model-layer speed.

Sources