July 25, 2026

AIincider

AI News. No Noise. Just Signal.

Claude Opus 5 Arrives: Flagship AI at Half the Price

2 min read
Anthropic launched Claude Opus 5, offering near-flagship performance at half the input price with a 1M-token context and reasoning control. Read more.

Anthropic has released Claude Opus 5, its most capable model yet, and the headline is not the intelligence but the invoice. The new flagship reaches roughly Fable 5 level performance for half the input price. It landed on July 24, 2026, making it Anthropic’s fourth new model in under two months.

From premium tier to value play

Opus has always sat at the top of Anthropic’s lineup, reserved for the hardest reasoning, coding, and agentic tasks and priced to match. The company spent June rolling out Mythos 5, Fable 5, and Sonnet 5, steadily pushing capability down the cost curve. Opus 5 continues that push at the high end, delivering near top tier results without the top tier bill.

What Anthropic announced

Claude Opus 5 costs $5 per million input tokens and $25 per million output tokens, the same rate as Opus 4.8 and half the input price of Fable 5. A separate fast mode doubles the rate to $10 and $50 while running about 2.5 times quicker. Cached input is billed at one tenth of the base rate, and asynchronous batch jobs get a 50 percent discount.

The model ships with a 1 million token context window and a new reasoning control that can be set per request to low, medium, or high. Low mode is quick and cheap for routine work, while high lets the model think longer on difficult problems. Opus 5 is now the default model on Claude Max and is available in the API as claude-opus-5.

Why it matters

The pricing is the story. On Anthropic’s internal Frontier Bench v0.1, Opus 5 scored 43.3 percent against Fable 5’s 33.7 percent, with gains concentrated in agentic coding and computer use. Offering that jump at half the input cost puts direct pressure on rivals: OpenAI’s GPT-5.6 family and Google’s delayed Gemini 3.5 Pro now face a cheaper frontier option for developers building agents.

For teams weighing cost against capability, the calculus just shifted. Watch how quickly competitors respond on price, and whether the low, medium, and high reasoning dial becomes a standard feature across the industry.

Continue Reading…

Leave a Reply