Anthropic Releases Claude Opus 5: Same Price, Higher Capability
Anthropic announced Claude Opus 5, its most capable model, and made it available across all platforms the same day. The notable part is that the price didn't move: it costs the same as the previous Opus 4.8, yet the company says capability is meaningfully higher.
In brief
- What shipped: Claude Opus 5 (model ID
claude-opus-5), released July 24, 2026. It's available on the Claude API, Claude.ai, Claude Code, and Claude Cowork; it's the new default on Claude Max and the strongest option on Claude Pro. - Price: $5 input / $25 output per million tokens — identical to Opus 4.8. There's also a "Fast mode" that runs 2.5× faster at 2× the base price.
- How much better: By Anthropic's own measurements, it's a big step over Opus 4.8 — it doubles its score on Frontier-Bench and scores "three times the next-best model" on ARC-AGI 3. In life sciences it reports gains of 10.2 points in organic chemistry and 7.7 in protein tasks.
- Safety: Based on an automated behavioral audit, the company calls it its "most aligned model to date."
Our take
The real story here isn't "a new model shipped" — that happens monthly — it's that the price held flat. A more capable model at the same budget is a direct win for anyone paying per token or on Pro/Max: your tool gets stronger without you doing anything. Practical tip: if you're on Pro or Max you've already been moved to the new model, so there's no need to manually pick the old default before starting a task. One honest caveat: these benchmark figures are the company's own tests; independent comparisons take a few weeks, so it's early to call it "best at everything." Still, adding capability without raising the price is a concrete, positive move on the user's side.
Kaynak: Anthropic