// the IDE for humans and AI.
px0 is the IDE for humans and AI, optimized for quick, fast code reviews. With native Git and GitHub integrations, open any PR URL directly, review scoped merge-base diffs, draft inline comments, and commit or push fixes back in place. Connected directly to your local coding harnesses (Claude Code, Gemini, Cursor, and more), px0 turns code reviews and agent edits into a zero-latency loop.
Built from the ground up for speed: instant PR checkouts, native Git operations, and seamless handoff to your AI coding harnesses.
Paste any GitHub PR URL to review scoped merge-base diffs, draft line comments with Alt+R, and submit formal approvals or batch changes without waiting for heavy IDEs.
$ px0 https://github.com/owner/repo/pull/123 → opened PR #123 in 12 ms · ephemeral worktree ready $ # drafted 2 inline comments · Alt+R ✓ submitted review with Approve to GitHub
Select code in px0 or review comments and dispatch directly to Claude Code, Gemini CLI, Cursor Agent, Antigravity, OpenCode, Codex, Aider, or Goose with live reloading.
px0: Batch Apply (2 review comments) → dispatched to claude-haiku runner ✓ 3 files refactored · tabs reloaded in place git panel: Commit with AI & push to PR branch
Read through massive, 50,000+ file codebases on a laptop without background indexers hogging RAM or spinning up fans.
$ px0 ~/src/linux indexed 95,710 files in 370 ms · RSS 20.4 MB $ # fuzzy search in 6 ms · CPU < 1% ✓ smooth scrolling across million-line trees
Spin up on any remote server, VM, or CI runner and browse the codebase instantly from your local browser: no SSH keys, no port forwarding hassle, and no heavy remote desktop daemons.
runner@ci-box:~/repo$ px0 -host 0.0.0.0 -no-open listening on http://10.0.4.12:7777 local-laptop:~$ open http://dev-runner:7777 ✓ connected directly from local browser
Every number here comes from ./benchmark.sh against shallow clones of seven real repositories. Linux, language servers off, fastest of five requests. Run it on your machine and paste the table anywhere.
| editor | process arch | memory (rss) | time to open | first interact |
|---|---|---|---|---|
| px0 | 1 process (native Go) | ~15 - 18 MB | ~10 ms | ~15 ms |
| Vim | 1 process | ~10 - 15 MB | ~15 ms | ~15 ms |
| Neovim | 1 process | ~10 - 20 MB | ~150 ms | ~150 ms |
| Sublime Text | 2-4 processes (C++) | ~100 - 250 MB | GUI dep. | ~250 - 500 ms |
| Zed | 1-3 processes (Rust) | ~200 - 450 MB | GUI dep. | ~300 - 600 ms |
| VS Code | 12 - 15+ processes | ~1,100 - 1,440 MB | ~3.0 - 5.0 s | ~6.0 - 10.0 s |
| repo | lang | source | files | index | fuzzy | full scan | open big | reopen | rss | peak |
|---|---|---|---|---|---|---|---|---|---|---|
| flask | Python | 3 MB | 235 | 1 ms | 0.8 ms | 2.3 ms | n/a | n/a | 16 MB | 18 MB |
| redis | C | 26 MB | 1,855 | 13 ms | 1.0 ms | 18.2 ms | 80.8 ms | 1.2 ms | 17 MB | 27 MB |
| django | Python | 74 MB | 7,014 | 39 ms | 1.3 ms | 26.8 ms | 166.8 ms | 1.0 ms | 20 MB | 29 MB |
| react | JavaScript | 63 MB | 7,178 | 52 ms | 2.7 ms | 32.2 ms | 57.7 ms | 0.7 ms | 21 MB | 28 MB |
| kubernetes | Go | 370 MB | 25,926 | 150 ms | 13.5 ms | 84.6 ms | 199.0 ms | 0.9 ms | 30 MB | 44 MB |
| typescript | TypeScript | 414 MB | 66,533 | 566 ms | 6.2 ms | 150.3 ms | 40.9 ms | 6.5 ms | 69 MB | 105 MB |
| linux | C | 1,809 MB | 95,710 | 370 ms | 6.0 ms | 451.8 ms | 26.7 ms | 0.6 ms | 55 MB | 73 MB |
px0 reclaims memory after 15 s of inactivity, so an open session settles back down.
Servers are separate processes, spawned on the first request for their language and shut down on exit. Without one, px0 falls back to a regex outline instantly.
$ ./benchmark.sh --clone && ./benchmark.shmethodology → BENCHMARKS.mdThe code review, Git inspection, and AI harness tools you reach for a hundred times a day, done in milliseconds, in a UI that renders only the lines you can see.
Files, symbols, and a full-tree regex scan behind three chords. Ranking favours basenames, word boundaries, and camelCase humps, the way fzf feels.
Pass any PR URL to review merge-base diffs, draft inline comments (Alt+R), and run 1-click batch AI edits. Stage per-file ticks, commit with AI, and push straight back to PR branches.
Stage multiple selection comments across files with Alt+E, review unified batch controls, and apply all edits at once with Mod+Enter. Live Git status sync highlights touched tabs instantly.
Full Shift and Ctrl keyboard selection across lines, words, and pages. Select code inside side-by-side or unified diffs, right-click for instant actions, or copy standardized @path:l1-l2 snippets.
Full GitHub Flavored Markdown with syntax-highlighted code fences and bi-directional scroll sync. Inspect images with zoom/pan, contrast modes (b), pixelated toggle (p), and lightbox.
Windowed tokenizing with Chroma...
Opening a 400,000-line file costs the same as a 10-line one. Only the visible window plus 24 rows of overscan is mounted; scrolling recycles rows with a single transform.
Native tokenizing through Chroma. The server lexes a 1,000-line window padded with 400 lines of context, never the whole file, and caches rendered HTML under a 512 MB LRU budget.
Zero-config detection on $PATH and standard binary folders. Spawned lazily on first request, stopped on exit. Click the status bar hint for automatic installation recipes.
Spin up on cloud VMs, Docker containers, or CI runners with -host 0.0.0.0. Browse in your local browser over Tailscale or WireGuard with zero remote extension daemons and strict DNS rebinding checks.
$ px0 -host 0.0.0.0 -port 7777 ~/work listening on: http://10.0.4.12:7777 sandboxing: path traversal + DNS guards
Tokyo Night by default, plus Paper, Catppuccin, Dracula, GitHub Dark, Gruvbox, Monokai, Nord, One Dark, Rose Pine, and Solarized. A theme is one CSS file of tokens.
HTML, CSS, and JS are embedded with go:embed. No Electron, no Node, no CGO, no system libraries. Copy it to a box and run it.
px0 static, go:embed ├─ index · search · highlight · lsp · agent ├─ web/ html css js, ~74 KB └─ themes/ 14 × css
After 15 seconds without requests, px0 automatically frees memory back to the kernel. A tab left open for the afternoon costs what an idle one should.
Spin up on any remote server, cloud instance, or runner in under a millisecond. Inspect and search remote codebases directly in your local browser without configuring SSH keys, forwarding ports through flaky sessions, or installing heavy remote desktop suites.
$ px0 -host 0.0.0.0 -port 7777 ~/work/repo$ px0 -no-open -port 8080 /workspace$ docker run -p 7777:7777 -v $(pwd):/src px0:latestRemote code editors install gigabytes of node runtimes and background server processes on the remote machine. px0 runs as a single lightweight binary, serving directly over HTTP without requiring you to maintain SSH shell sessions.
Pass -host 0.0.0.0 or -port 7777 to bind the web console. Access from any browser on your laptop, iPad, or phone across your private VPN, Tailscale mesh, or Cloudflare Tunnel with pooled gzip compression.
Exploration is strictly sandboxed: safePath rejects directory traversal attempts, host header checks protect against DNS rebinding, and external stdlib jumps require explicit allowlisting. Agent editing is protected by origin verification: requests are accepted only when opened by IP address or localhost, and strictly refused through public tunnel hostnames.
Pass any pull request URL directly to px0. Inspect merge-base diffs, draft inline comments, run one-click batch AI edits, stage per-file ticks, commit with AI, and push fixes straight back to the PR's own branch without leaving your browser.
Apply all drafted review comments across files with your local coding agent.
Pass any pull request URL to inspect diffs in an isolated ephemeral worktree. Full codebase navigation, symbols, definitions, and callers remain accessible with zero disk clutter.
Hover any line or press Alt+R to draft inline review notes. Submit formal approvals to GitHub or click Batch Apply to dispatch all comments across the PR to your local AI coding agent.
Stage per-file ticks in the tree, generate commit messages from staged diffs with AI, and push directly back to the PR head branch (including forks) without opening a terminal.
Code now arrives from the terminal, not the keyboard. What is left for you is review, audit, and navigation, and that job needs one thing done exceptionally well: a zero-latency, distraction-free place to read, explore, and verify. Not an editor that takes seconds to boot and gigabytes to idle.
drives generation
Agents, harnesses, and orchestrators write the code, run the tests, and move the workflow along.
agent 14 files changed ✓ 212 tests passed
drive verification
Follow symbol references, inspect git diffs, preview rendered docs, and check the architecture.
Ctrl D diff vs HEAD Alt M preview F12 definition Shift F12 refs
is overkill
Booting an Electron suite with an extension host and a gigabyte-plus heap just to read agent output.
1,440 MB rss, 15+ processes 4 - 10 s to index the workspace
| dimension | traditional IDE | px0 |
|---|---|---|
| built for | typing code character by character, hosting plugins | humans and AI: reading, exploring, diffing, and verifying code |
| base memory (rss) | ~1,440 MB | ~20 MB70x lighter |
| startup cpu spike | 35% - 50% | < 1% |
| cold start | several seconds | sub-millisecond |
| process tree | 15+ Node.js / Electron processes | 1 static Go binary |
| workspace indexing | multi-second background churn | 13 ms for redis, 370 ms for the linux kernel |
| setup | config files, plugins, node, npm | zero config, zero runtime |
The chords you already know from your editor, plus one palette that switches mode on its first character. Press ? inside px0 for the full sheet.
lspsrv→ lspservers.go>diff→ Toggle Git Diff@score→ fuzzyScore:120→ go to line 120A Go server that owns indexing, search, highlighting, and language servers, talking JSON to a zero-dependency ES module frontend. The speed comes from a pile of small, boring decisions.
Pass one proves every query rune is present. Pass two walks back from the last hit to pull matches as tight as possible. The same ranking feel as a dynamic program, in a linear scan. ARCHITECTURE.md →
If your engineers are actively using coding agents and looking for a fast, ultra-lightweight code reader and diff viewer to speed up verification, we would love to collaborate.
When agents generate large diffs across dozens of files, inspection speed is key. px0 starts in under 1ms with ~20 MB RAM, giving your team instant fuzzy search and split diffs.
Get a direct line with the core team on Slack or Slack Connect. We prioritize bug fixes, performance tuning, and integration with your specific agent toolchains.
px0 runs strictly locally on developer laptops or remote dev boxes. Zero telemetry, zero cloud dependencies, and zero risk of proprietary code leaking.
No results matching ""
Try searching for keywords like docker, flags, keys, sudo, or lsp.