Anthropic launched an AI code reviewer. Some developers say it's expensive and undermines senior engineers.

• Anthropic released Code Review, which is designed to catch complex coding issues and fix bugs.
• As the feature "optimizes for depth," the company said it "is more expensive than lighter-weight solutions."
• Developers on X said the tool is costly and could chip away at senior engineers' roles

Anthropic gave Claude a new job: reviewing code.

But some developers say the new feature could chip away at senior engineers' roles and inflate token bills.

The AI startup announced in a blog post on Monday that Claude's Code Review is designed to catch more complex issues and suggest fixes for bugs. Since the feature "optimizes for depth," Anthropic said it "is more expensive than lighter-weight solutions like the Claude Code GitHub Action."

Claude Code GitHub Action is an integration that lets developers use Claude to review code in GitHub, a platform for storing and collaborating on code.

"Reviews are billed on token usage and generally average $15—25, scaling with PR size and complexity," the company added, referring to pull requests — a common workflow where developers propose changes to a codebase.

An Anthropic engineer who works on Claude Code, Thariq Shihipar, said on X on Tuesday that the feature "uses a lot more compute and tends to catch more difficult bugs."

Shihipar was responding to criticism from another developer, Daniel Ávila, who goes by Daniel San online and describes himself as head of AI at the fintech startup Hedgineer.

Ávila said in an X post on Tuesday that he saw "no improvement" using Code Review.

"Other than not needing a trigger like a GitHub Action and being configurable directly inside Claude Desktop, I see absolutely no additional functionality," Ávila said.

"The only real difference is that GitHub Actions uses the API and Code Review doesn't," he added.

Reactions among developers have been mixed, particularly around the price.

Jean P.D. Meijer, a software engineer and founder of a software startup, questioned the pricing in a post on X.

"How is Claude Code Review $15 to $25 per PR?!" Meijer wrote under the handle @initjean.

"One review is basically the price of an unlimited sub to most code review apps," he added. "Did they forget to subsidize the tokens?"

Another developer at a design startup, Karan Kendre, also questioned the logic behind the feature.

"If Claude Code is capable of working clean code in a codebase why can't he write write it from the start," Kendre wrote on X on Tuesday.

"Because they want to charge 15-25$ per code review on top of the claude code subscription," he added.

A developer who goes by @TukiFromKL and is building an AI startup wrote that "Claude just put a price tag on your senior developer."

"$15—25 per code review. AI agents that start instantly the second you open a PR," the user wrote.

"The people protecting their $200K salaries by gatekeeping code reviews are about to have a very bad quarter," the user added.

AI coding tools are getting expensive

Anthropic isn't the only AI company facing scrutiny over token costs.

Billionaire investor Chamath Palihapitiya said in an episode of the "All-In Podcast" published Friday that his software startup is reconsidering its use of Cursor because of rapidly rising expenses.

"Our costs have more than tripled since November," he said. "Between the inference cost that we pay AWS, which is ginormous, between our cost with Cursor, between Anthropic, we are just spending millions."

Palihapitiya said his startup, 8090, is on track to spend roughly $10 million a year on AI costs.

"The problem is that my costs are going up three times every three months," he said. "My revenues are not."

He pointed to Cursor as one contributor to the company's growing bill, and said Claude Code has been comparatively cheaper.