Anthropic has released Claude Opus 5, its newest AI model, which the company says comes close to the capabilities of its flagship Claude Fable 5 at roughly half the price. According to Anthropic, Opus 5 is available now and becomes the new default model on Claude Max. The company also called it its "most aligned model to date," as reported by Inshorts.
The pitch is economics as much as raw power. CNBC reports Anthropic describes Opus 5 as both its best-performing and most cost-effective offering, and Bloomberg frames it as a more cost-efficient model built for everyday tasks. Reuters, via Livemint, characterized the launch as an efficiency upgrade aimed at daily work. According to ZDNET's David Gewirtz, Opus 5 is priced at $5 per million input tokens and $25 per million output tokens — the same as the earlier Opus 4.8 — while improving coding, reasoning efficiency and tool use. The model targets coding, knowledge work and agent workflows.
On benchmarks, R&D World reports Opus 5 outscores Fable 5 on 8 of 13 tests at half the token price, and PCMag says Anthropic claims it tops Fable 5 on agentic search. Opus 5 was listed at #1 on the Artificial Analysis Intelligence Leaderboard, according to a Hacker News post. Fortune notes a feature letting users toggle between cost and capability. Not everyone is convinced: Yellow.com reports experts are split after the first independent tests.
The launch lands amid notable timing. The Verge notes it came days after an OpenAI security incident and weeks after Anthropic's latest dealings with the U.S. government, while Yahoo Finance reports Anthropic is preparing for an IPO later this year. Snowflake has already integrated the model, per StartupHub.
Why it matters: as businesses worry about the rising cost of AI, Anthropic is betting that cheaper models offering near-flagship performance — not just more powerful ones — are what will win over enterprise customers.