Key Takeaways:
- Opus 5 costs $5 per million input tokens, half Fable 5's price
- It scores 43.3% on Frontier-Bench, more than double Opus 4.8's 18.7%
- Anthropic positions it as the daily driver for enterprise AI workloads
뒤로
Anthropic's Opus 5 delivers near-frontier AI at half the cost
Anthropic released Claude Opus 5 on July 24, positioning it as a near-frontier model that costs half as much as its top-tier Fable 5 — a bet that enterprise AI spending will shift toward efficiency over raw capability.
The new model, priced at $5 per million input tokens and $25 per million output tokens — unchanged from its predecessor Opus 4.8 — becomes the default on Anthropic's Max subscription and the strongest option on Pro. The company is not claiming Opus 5 is its smartest model; that distinction still belongs to Fable 5, which costs roughly twice as much. Instead, Anthropic is making a subtler argument: that the most economically important AI work happens in a middle band of difficulty, where near-frontier intelligence delivered efficiently beats frontier intelligence delivered expensively.
"Opus 5 as your daily driver, the model you hand complex work to and review when it's done," an Anthropic spokesperson told VentureBeat. "Fable 5 for your most ambitious work, the days-long autonomous projects nothing could take on before."
On benchmark tests, Opus 5 sets new highs on coding and knowledge-work evaluations including Frontier-Bench and GDPval-AA. It scores 43.3 percent on Frontier-Bench v0.1, an agentic terminal coding benchmark — more than double Opus 4.8's 18.7 percent and ahead of Fable 5's 33.7 percent, at a lower cost per task. On ARC-AGI 3, a test of novel problem-solving, Anthropic says Opus 5 scored three times as high as the next best model. On OSWorld 2.0, a computer-use benchmark, it beats Fable 5's best result at just over a third of the cost. On CursorBench, it lands within 0.5 percent of Fable 5's peak at maximum effort while costing half as much per task.
The company acknowledges Opus 5 remains behind Mythos 5, a competing model, on cybersecurity tasks and biology research. In one internal biology exercise, the model repeatedly got stuck in self-verification loops and failed to deliver protein designs that Mythos 5 completed. Anthropic also notes that Opus 5 "hallucinates factual claims slightly more than Opus 4.8, despite being more accurate overall."
Customers report fewer tokens, fewer turns, less time
Early adopters described efficiency gains that go beyond benchmark scores. Harvey, the legal AI company, said Opus 5 achieved similar performance to Opus 4.8's maximum-reasoning mode "while generating 26 percent fewer tokens on average," according to Niko Grupen, its head of applied research. Richard Pham of Fundamental Research Lab said that on hard financial-modeling tasks, the model averaged nine percentage points higher accuracy "while using roughly one-third fewer turns and tool calls and 60 percent less time."
Wade Foster, chief executive of Zapier, said Opus 5 topped his company's AutomationBench leaderboard "without spending more tokens than prior Claude models," running a full churn-prevention workflow from start to finish. "Previous models didn't pass; Opus 5 hit 100 percent," he said. Scott Wu, chief executive of Cognition, the company behind the Devin coding agent, said that on FrontierCode 1.1, "Claude Opus 5 approaches Fable-level performance at half the cost," with particular strength in debugging and root-cause analysis.
The model ships with an adjustable effort setting that lets customers trade reasoning depth for lower token cost and faster responses, plus a Fast mode running at roughly 2.5 times default speed at twice the base price. Anthropic also added automatic fallback routing on the API and mid-conversation tool changes that no longer invalidate the prompt cache — a feature agent developers have requested.
Safety by design, with deliberate capability gaps
Anthropic says Opus 5 is its most aligned model to date, scoring 2.3 on its internal misaligned behavior audit — lower than Opus 4.8, Sonnet 5, or Fable 5 — with the lowest rates of deceptive behavior and susceptibility to being tricked into misuse. The company intentionally avoided training Opus 5 on cyber tasks, as it did with Opus 4.8, though the model improved on them anyway as a side effect of general capability gains. On Anthropic's OSS-Fuzz evaluation, Opus 5 identified vulnerabilities at a 79.4 percent rate, close to Mythos 5's 80 percent, but succeeded at developing exploits in only four challenges versus Mythos 5's 13 — an asymmetry that appears deliberate.
When safety classifiers do trigger, requests fall back to Opus 4.8 by default. "The model it falls back to has lower capability levels making the risk of harmful use lower as well," the spokesperson said. Unlike Fable and Mythos, Opus 5 carries no data retention requirement for general access — a point the company flagged for customers with "a hard zero data retention requirement."
The launch comes at a moment of extraordinary commercial momentum for Anthropic. The company was valued at roughly $380 billion in its latest funding round, according to Reuters, following a period in which annualized revenue climbed from about $1 billion at the end of 2024 to a projected $9 billion by the end of 2025, with internal targets reportedly reaching $20 billion to $26 billion for 2026. Those targets are underwritten by enormous infrastructure commitments, including a reported $30 billion Azure compute deal alongside arrangements with Google Cloud and Nvidia.
For investors, the implications cut across the AI value chain. Anthropic's pricing strategy — holding Opus 5 at Opus 4.8 levels while roughly doubling performance on key agentic benchmarks — is effectively a steep price cut per unit of capability, designed to widen the funnel of workloads that are economical to automate. Every task that was marginal at Opus 4.8's cost-per-success becomes viable at Opus 5's, and every viable task is recurring token revenue. The question for Nvidia, Microsoft, and Google is whether this efficiency push expands the total addressable market enough to offset lower revenue per inference — or whether it signals a market where model commoditization compresses margins for infrastructure providers. Anthropic's bet is that cheaper intelligence creates more demand, not less.
This article is for informational purposes only and does not constitute investment advice.
