Daily Delta · the full list

4 AI tools that review your code before a human does

Thanks for commenting "REVIEW" — here's everything from the post in one place: what each tool actually does, its real free tier, and where it falls short.

Went through free-for-dev — the internet's biggest curated list of free dev tools, built by 1,600+ contributors — looking for what's genuinely AI-powered in code review, not just "automated." Only 4 tools on the entire list qualified. Most of what gets labeled "automated code review" out there (Codacy, CodeFactor, DeepSource, sonarcloud.io, and others on the same list) is static analysis with no AI in the loop — useful, but a different category.

01 · CodeRabbit

Reviews every pull request automatically

CodeRabbit reads every PR, summarizes what changed, flags likely bugs, and suggests one-click fixes directly in the review thread.

Key facts

Free tierFree forever for public/open-source repos. Private repos get a 14-day full-featured trial, then fall back to summaries only.
PaidPro $24/user/mo, Pro Plus $48/user/mo, Enterprise custom
IntegratesGitHub, GitLab, Azure DevOps, Bitbucket, VS Code, Cursor, Windsurf, CLI, Jira/Linear, MCP

CodeRabbit cites "6M repositories analyzed, 75M defects found" on its site — the vendor's own number, not independently verified here.

02 · What The Diff

Writes your PR descriptions for you

What The Diff reads the diff and writes the pull request description, changelog entries, and a plain-language summary for non-technical stakeholders — so reviewers get context without the author having to write it.

Key facts

Free tier25,000 tokens/month (~10 PRs), no card required
Paid$19/mo for 200,000 tokens (~40 PRs) — unused tokens don't roll over
IntegratesGitHub and GitLab only

03 · Pullflow

Syncs AI and human review status in one place

Pullflow isn't itself an AI reviewer — it's a collaboration layer that pulls other tools' AI reviews (GitHub Copilot, CodeRabbit, Greptile) plus human review status into one shared view across GitHub, Slack, and VS Code.

Key facts

Free tierUnlimited public repos, up to 5 private-repo users, forever, no card
PaidTeam $5/user/mo (annual), Enterprise custom
IntegratesGitHub (repos + Actions/CI), Slack, VS Code (incl. Cursor, Antigravity)

Pullflow cites "~15% of PRs had AI agent participation in 2025, up from ~2.6% in 2024" and "merge PRs 4x faster" — both vendor-cited figures, not independently verified here.

04 · Zenable

Enforces your team's standards on AI-written code

Zenable combines static analysis with AI review to catch security issues and enforce your team's own coding standards on AI-generated output (Cursor, Copilot, Claude Code) before it merges.

Key facts

Free tier15 PR reviews/week (shared team pool) + 100 agentic/MCP reviews per day (per seat)
IntegratesGitHub App, GitLab, 50+ IDEs, Claude Code, MCP server, CLI, git hooks
NoteStates on-site it does not train on your code

Recap

CodeRabbit — reviews every PR, free forever for public repos
What The Diff — auto-writes PR descriptions, ~10 PRs/month free
Pullflow — syncs AI + human review status across GitHub and Slack
Zenable — enforces your team's own standards on AI-written code

The catch

None of these replace a senior engineer's judgment on whether a feature is actually right. They catch bugs, style issues, and drift from your standards — not bad product decisions. Use them to clear the small stuff so human reviewers can spend their time on the part that actually needs a human.

Links

CodeRabbit What The Diff Pullflow Zenable free-for-dev (source list)