# Anthropic Launches Claude Opus 5 as a Cheaper Coding and Agent Model

**Anthropic's new Opus 5 model costs the same as its predecessor but beats Fable 5 on key coding benchmarks.**

## What Anthropic Announced

Anthropic released Claude Opus 5 on Friday, positioning the model as nearly as capable as its top-tier Claude Fable 5 at roughly half the running cost. The model went live immediately across Anthropic's platforms and became the default model on Claude Max, the company's premium consumer tier, while also standing as the strongest option available on Claude Pro. Rather than calling Opus 5 its smartest model overall — a title the company still reserves for Fable 5 — Anthropic is framing the release around a different question: which model gets ordinary, high-volume work done most cheaply? An Anthropic spokesperson told [VentureBeat](https://venturebeat.com/orchestration/anthropic-launches-claude-opus-5-a-cheaper-ai-model-for-coding-agents-and-enterprise-workflows) that Opus 5 is meant as a daily driver, the model teams hand complex work to and review once it is finished, while Fable 5 stays reserved for the most ambitious, days-long autonomous projects.

## Pricing and Positioning

Opus 5 is priced at $5 per million input tokens and $25 per million output tokens, unchanged from its predecessor, Opus 4.8. That pricing stability matters because Anthropic is betting the bigger shift is efficiency, not sticker price: the company argues that the most economically valuable AI work sits in a middle band of difficulty, where near-frontier intelligence delivered cheaply beats frontier intelligence delivered expensively. The company's lineup now stratifies by task type, with Sonnet 5 handling large-scale work where speed and per-call cost decide what ships.

## Benchmark Results

On the Frontier-Bench coding evaluation, Anthropic reports Opus 5 scoring 43.3 percent, ahead of Fable 5's 33.7 percent and more than double Opus 4.8's 18.7 percent, all at a lower cost per task. On the OSWorld computer-use benchmark, the company says Opus 5 surpasses Fable 5's best result at just over a third of the cost. Anthropic is candid about the gaps too: Opus 5 still trails a rival model called Mythos 5 on cybersecurity tasks and biology research, and an OpenAI-family model keeps its lead on one agentic coding benchmark. The company's own explanation is that Opus 5 wins bounded tasks with a clear finish line, while Fable 5 is built for jobs that stretch across many connected steps over hours or days.

## What Enterprise Customers Are Saying

Early customers backed up the efficiency claim with specific numbers. Harvey, a legal AI company, said Opus 5 matched Opus 4.8's top reasoning mode while generating 26% fewer tokens on average. Fundamental Research Lab reported nine percentage points higher accuracy on financial-modeling tasks while using roughly one-third fewer turns and tool calls and 60% less time. Zapier's chief executive, Wade Foster, said Opus 5 topped the company's AutomationBench leaderboard without spending more tokens than prior Claude models, completing a full churn-prevention workflow that earlier models could not finish at all — "Opus 5 hit 100%," he said. Cognition's chief executive, Scott Wu, said the model approaches Fable-level performance at half the cost on coding tasks, with particular strength in debugging and root-cause analysis. Anthropic's business skews heavily toward enterprise usage: Claude holds roughly 40 percent of the enterprise large language model market by usage, and Claude Code alone has reached about $1 billion in annualized revenue.

## Key facts

- Opus 5 is priced at $5 per million input tokens and $25 per million output tokens, the same as Opus 4.8.
- On Frontier-Bench, Opus 5 scored 43.3 percent versus Fable 5's 33.7 percent and Opus 4.8's 18.7 percent.
- Harvey reported 26% fewer tokens for similar performance; Zapier reported 100% completion on its AutomationBench workflow.
- Claude holds roughly 40 percent of the enterprise large language model market, and Claude Code has reached about $1 billion in annualized revenue.

## Check your understanding

What price does Anthropic charge per million input and output tokens for Claude Opus 5?
How did Opus 5's Frontier-Bench score compare with Fable 5 and Opus 4.8?
Why does Anthropic say Fable 5 is still better suited to long-horizon autonomous projects than Opus 5?

## Sources

- [VentureBeat: Anthropic launches Claude Opus 5](https://venturebeat.com/orchestration/anthropic-launches-claude-opus-5-a-cheaper-ai-model-for-coding-agents-and-enterprise-workflows)

---

- Category: know (release)
- Affects: agent builders · enterprise spend
- Published: 2026-07-24T17:00:00.000Z
- Original: https://venturebeat.com/orchestration/anthropic-launches-claude-opus-5-a-cheaper-ai-model-for-coding-agents-and-enterprise-workflows
- Generated by tinker news