AI Models, explained & compared

For most people, Claude Opus 5 — released July 2026, it approaches the frontier intelligence of Fable 5 at half the price ($5 vs $10 per million input tokens). Reach for Claude Fable 5 only when you need the absolute ceiling on long-horizon agentic work, and drop to Sonnet 5 or Haiku 4.5 when speed and cost matter more than raw capability.

Fable 5 is Anthropic's most capable model and its priciest at $10/$50 per million tokens. Opus 5 lands just behind it on most benchmarks — within 0.5% on coding evaluations like CursorBench — but costs half as much and runs faster. Unless you are pushing the most demanding autonomous agents, Opus 5 is the better value.

Many do. Cursor, Bolt and other agentic coding tools let you pick Claude as the model behind the assistant, and several builders use it by default. The model you choose changes how well the tool reasons about your code — which is why understanding the Claude lineup helps you get more out of the builders we review.

API pricing per million tokens (input / output): Haiku 4.5 is $1/$5, Sonnet 5 is $3/$15, Opus 5 is $5/$25, and Fable 5 is $10/$50. Chat access through Claude Pro and Claude Max bundles the top models into a flat monthly subscription instead of per-token billing.

Yes — at the same price. Anthropic reports that Opus 5 roughly doubles Opus 4.8's score on its Frontier-Bench coding benchmark and posts double-digit gains on scientific tasks, all at the identical $5/$25 per-million-token pricing.