Opus 5 halves Anthropic's top-tier price while beating Fable 5 on most tests

Anthropic has cut the price of its top-tier intelligence in half. Claude Opus 5, launched 24 July, matches Claude Opus 4.8's pricing of $5 per million input tokens and $25 per million output tokens, undercutting Claude Fable 5 by roughly 50% while beating it on 8 of 13 benchmarks. For buyers choosing between Anthropic's own models, that makes Fable 5 a harder sell on pure economics.

Opus 5 is now the default model for Claude Max subscribers and the strongest available on Claude Pro. It ships on Claude, Claude Code, the Claude API under the model ID claude-opus-5, and across the three major clouds: Amazon Bedrock, Google Cloud Vertex AI and Microsoft Foundry, according to SiliconANGLE.

The model carries a 1-million-token context window and up to 128,000 output tokens, plus a new low/medium/high effort setting that lets developers trade cost for capability on a per-request basis. On Frontier-Bench, a set of 74 physics, chemistry and cryptography tasks, Opus 5 scored 43.3%, 9.7 percentage points above Fable 5. It also beat Fable 5 by 8.5 points on AutomationBench.

Anthropic is pitching Opus 5 as its safest release yet. It scored zero on the company's internal benchmark for exploiting vulnerabilities, can scan code for security issues while blocking methods attackers commonly use, and triggers safety guardrails 85% less often than Fable 5. New API controls let developers interrupt a model mid-workflow to swap its tools, and reroute prompts blocked by safety filters to a different model instead of failing outright.

The launch is Anthropic's fourth model release in under two months, following Mythos 5, Fable 5 and Sonnet 5 in June. A checkpoint codenamed "Claude Honeycomb EAP" briefly surfaced in Cursor's model picker on 9 July, ahead of the official announcement.

Halving the price of its flagship tier while improving benchmark performance is an aggressive move against Anthropic's own recent release, not just rivals. Enterprises that adopted Fable 5 weeks ago now face a cheaper, stronger in-house alternative, which pressures Anthropic to justify Fable 5's continued existence and gives procurement teams leverage to renegotiate.

The cadence is the bigger signal: four models in under two months suggests Anthropic is optimising for rapid iteration over long release cycles, a strategy that keeps competitors reacting to a moving target but also risks fragmenting its own product line. Watch whether Fable 5 gets repriced or quietly retired.