Anthropic released Claude Opus 5 on July 24, 2026 — its fourth model launch in under two months, following Mythos 5, Fable 5, and Sonnet 5 in June. Rather than leading with a new capability ceiling, Anthropic framed the release around economics: a model built to be used every day rather than saved for the hardest problems.
Opus 5 is priced identically to its predecessor, Opus 4.8, at $5 per million input tokens and $25 per million output tokens — exactly half of what Anthropic charges for its flagship Fable 5 model, according to TechCrunch’s coverage.
Despite the lower price tier, Opus 5 actually outperforms Fable 5 on several of the benchmarks Anthropic published at launch — an unusual position for a company’s second-tier model to be in.
The numbers that matter for buyers
On Frontier-Bench v0.1, a measure of agentic coding ability, Opus 5 scored 43.3% — more than double Opus 4.8’s 21.1%, and roughly 10 points ahead of the next-closest model.
On ARC-AGI-3, a test of novel reasoning, it posted 30.2%, about three times the next-best score. Anthropic also introduced a low/medium/high effort toggle, letting developers dial reasoning compute up or down per request to manage cost directly.
The caveat worth knowing
Anthropic disclosed that its headline coding benchmark used a specific internal test harness where Opus 4.8 served as a fallback model on safety-classifier refusals for both Opus 5 and Fable 5 — meaning the reported scores aren’t a pure single-model result.
The methodology is well-documented by Anthropic, but it’s a vendor-run harness, and the 43.3% figure should be read with that context rather than taken as an independently verified number.
On CursorBench 3.2, a separate coding benchmark, Opus 5 at maximum effort landed within half a percentage point of Fable 5’s peak score — while costing roughly half as much per task. Anthropic also reported a lower misalignment score for Opus 5 than any model it has previously measured.
What this means for buyers
- Teams currently running Opus 4.8 for coding or agentic workflows get a meaningful capability jump at the same price point.
- Fable 5 remains the frontier option for the hardest tasks, but Opus 5 closes most of the gap at half the cost — worth testing before renewing enterprise Fable 5 contracts.
- The effort toggle gives finance and engineering teams a direct lever to manage AI spend without switching models entirely.
Sources
- Anthropic launches Opus 5 — TechCrunch
- Claude Opus 5: Frontier Intelligence at Half the Price — Digital Applied
- Claude Opus 5 Benchmarks Explained — Codersera
Disclaimer: This content is meant to inform and should not be considered financial advice. The views expressed in this article may include the author’s personal opinions and do not represent Times Tabloid’s opinion. Readers are advised to conduct thorough research before making any investment decisions. Any action taken by the reader is strictly at their own risk. Times Tabloid is not responsible for any financial losses.
Follow us on X, Facebook, Telegram, and Google News
Keywords
, , , , ARC-AGI-3, AI pricing, agentic coding, Claude Fable 5
