Download Latest Version Orca-1.4.243-arm64-mac.zip (273.7 MB) Google Add to Preferred Sources
Home / v1.4.241
Name Modified Size InfoDownloads / Week
Parent folder
latest.yml 2026-10-26 343 Bytes
orca-windows-setup.exe 2026-10-26 202.1 MB
orca-windows-setup.exe.blockmap 2026-10-26 202.7 kB
latest-linux.yml 2026-10-26 677 Bytes
latest-linux-arm64.yml 2026-10-26 690 Bytes
orca-ide-1.4.241.x86_64.rpm 2026-10-26 153.9 MB
orca-ide-1.4.241.aarch64.rpm 2026-10-26 149.4 MB
latest-mac.yml 2026-10-26 796 Bytes
orca-macos-arm64.dmg 2026-10-26 271.5 MB
orca-macos-arm64.dmg.blockmap 2026-10-26 285.1 kB
Orca-1.4.241-mac.zip 2026-10-26 281.5 MB
Orca-1.4.241-mac.zip.blockmap 2026-10-26 288.4 kB
Orca-1.4.241-arm64-mac.zip 2026-10-26 272.2 MB
Orca-1.4.241-arm64-mac.zip.blockmap 2026-10-26 280.9 kB
orca-macos-x64.dmg 2026-10-26 280.7 MB
orca-macos-x64.dmg.blockmap 2026-10-26 291.4 kB
orca-ide_1.4.241_amd64.deb 2026-10-26 181.8 MB
orca-ide_1.4.241_arm64.deb 2026-10-26 177.8 MB
orca-linux.AppImage 2026-10-26 244.8 MB
orca-linux-arm64.AppImage 2026-10-26 246.3 MB
orca-sourcemaps-v1.4.241.zip 2026-10-26 16.0 MB
README.md 2026-10-26 32.9 kB
v1.4.241 source code.tar.gz 2026-10-26 89.6 MB
v1.4.241 source code.zip 2026-10-26 113.5 MB
Totals: 24 Items   2.7 GB 32

Chats and a browser that behave the way you expect 🧭

This release makes chats follow your agent settings, brings Chrome-style tools to the built-in browser, and helps SSH connections and managed servers recover. You’ll also find a chat font setting, live agent browsing, and steadier Hermes, Kimi, Kiro and Windows terminals.

The short version

  • Chats that follow your settings: New chats use each agent’s permission from Settings → Agents, renamed chats keep their names in Session History, and a chat won’t write a conversation a terminal is still running.
  • A browser closer to Chrome: Dock DevTools inside the pane, mute tabs playing sound, see a link’s URL on hover, and use Orca shortcuts and mouse back/forward inside pages.
  • Watch agents browse: A cursor and badge show where an agent is clicking or typing, chat jumps you to the page, and links from terminals and chat open beside their source.
  • SSH with fewer dead ends: Saved port forwards return after a restart and explain failures, certificate-only hosts connect through your system ssh, and managed servers update without closing terminals.
  • Agents that start and resume reliably: Hermes and Kimi panes resume after a restart, Kiro reports status on its V3 engine, and keys typed while an agent starts are kept as a draft.
  • Polish where it counts: Pick a font for chat, zoom with Cmd/Ctrl + + on non-US keyboards, open cloud-only files from macOS terminals, and get steadier Windows terminals and pasting.

Phone & web updates: Phones no longer drop pieces of terminal output, and chat renames and each chat’s permission mode now show there. Session tabs also keep Working and Needs You for Windows agents that never set a title.

Compatibility & upgrade notes

  • Fresh terminal host: New terminals start on a fresh terminal host so this release’s terminal fixes apply. Terminals already open before you update keep running, and stay attached, on the previous host until you close them.
  • Chat permissions: Settings → Chat no longer has its own permission dropdown; set it per agent in Settings → Agents. Existing chats keep their saved permission, and a chat’s own permission menu changes only that chat.
  • Conversations open in a terminal: If a terminal might still be writing a conversation, its chat won’t send and asks you to quit the agent there or close that terminal. Closing that terminal lets the next send go through.
  • Managed SSH servers: Update anyway keeps running terminals going through the update. The offer to resume chats after a restart needs a server already running this release.
  • OpenCode model picks: OpenCode 2.0.14, 2.0.15, 2.0.17 and 2.0.18 leave the first message unsent when a model is picked, so Orca refuses model picks on those versions.

Product experience

Native chat (experimental)

Chats share your agent settings, keep their names, and stay out of conversations a terminal is still writing.

Browser

Chrome-style DevTools, tab audio and link previews, plus a live view of agents using the browser.

Agent launch, status & accounts

Agents start on the first try, report the right status, resume in the right place, and keep the right account.

Orchestration

Mail reaches workers when they can take it, and worker starts fail clearly instead of stalling.

Terminals

A fresh terminal host, steadier Windows terminals, and history and folders that survive sleep, restarts and phones.

SSH & remote servers

Port forwards that come back, more hosts that connect, and managed servers you can update without closing terminals.

Workspaces, sidebar & everyday polish

Workspaces stay on the right branch and dialogs don’t get stuck, with quicker previews and tidier sidebar controls.

Phone & web

Phones keep every piece of terminal output and the right agent status, and explain what they can’t show.

Tests, builds & maintenance

Broader behavioral coverage with fewer low-value tests, faster local checks and dev setup, continued module organization, a repository-wide comment cleanup, and relay service improvements.

All test, build and maintenance changes * Log the relay director’s zone and round-trip speed on each runtime line by @Jinwoo-H in https://github.com/stablyai/orca/pull/33372 * Cap how many reserve relay cells a director outage sends to the database by @Jinwoo-H in https://github.com/stablyai/orca/pull/33415 * Build the macOS terminal host on a patched Node runtime to recover stalled workers (reverted later in this release, apart from its stall telemetry) by @nwparker in https://github.com/stablyai/orca/pull/29133 * Return the macOS terminal host to official Node, keeping stall telemetry by @Jinwoo-H in https://github.com/stablyai/orca/pull/33688 * Clean up after the agent integration fixes by @nwparker in https://github.com/stablyai/orca/pull/33644 * Repair main’s line limits and a stale test import by @brennanb2025 in https://github.com/stablyai/orca/pull/33675 * Complete the repository-wide comment cleanup by @OrcaWin in https://github.com/stablyai/orca/pull/33249 * Finish the launcher and configuration comment review by @OrcaWin in https://github.com/stablyai/orca/pull/33513 * Fix suppressed lint violations instead of ignoring them by @OrcaWin in https://github.com/stablyai/orca/pull/33294 * Split STYLEGUIDE.md into a core guide plus topic deep dives by @nwparker in https://github.com/stablyai/orca/pull/33402 * Add component fixtures with headless screenshots for fast UI checks by @nwparker in https://github.com/stablyai/orca/pull/33406 * Give coding agents TypeScript 7 code intelligence, with load-aware type checkers by @nwparker in https://github.com/stablyai/orca/pull/33408 * Give each worktree its own pnpm dev profile, ports and seeded profiles by @nwparker in https://github.com/stablyai/orca/pull/33416 * Skip mobile dev-client rebuilds when the native fingerprint is unchanged by @nwparker in https://github.com/stablyai/orca/pull/33410 * Move mobile to TypeScript 7, Vitest 5 and the root typecheck runner by @nwparker in https://github.com/stablyai/orca/pull/33419 * Speed up the local test loop and changed-test runs in cold worktrees by @nwparker in https://github.com/stablyai/orca/pull/33421 * Keep JS plugins out of default oxlint and adopt ast-grep by @nwparker in https://github.com/stablyai/orca/pull/33430 * Make `pnpm check:changed` the single pre-handoff check by @nwparker in https://github.com/stablyai/orca/pull/33432 * Add an opt-in bundled renderer dev mode by @nwparker in https://github.com/stablyai/orca/pull/33433 * Serve the renderer with bundled dev mode by default by @nwparker in https://github.com/stablyai/orca/pull/33449 * Drop the load-based type checker cap by @nwparker in https://github.com/stablyai/orca/pull/33444 * Skip the mobile native fingerprint report on draft PRs by @nwparker in https://github.com/stablyai/orca/pull/33438 * Rerun a WSL2 CI job once when its hosted runner freezes by @nwparker in https://github.com/stablyai/orca/pull/33532 * Run topology SQLite fixture tests in Node by @nwparker in https://github.com/stablyai/orca/pull/33533 * refactor: group agent-session conversation and message identity by @nwparker in https://github.com/stablyai/orca/pull/33324 * refactor: group PTY delivery and drain ownership by @nwparker in https://github.com/stablyai/orca/pull/33358 * refactor: group mobile chat composer ownership by @nwparker in https://github.com/stablyai/orca/pull/33390 * refactor: group Parcel watcher process ownership by @nwparker in https://github.com/stablyai/orca/pull/33401 * refactor: group local and WSL filesystem watchers by @nwparker in https://github.com/stablyai/orca/pull/33407 * refactor: group filesystem IPC registration and suites by @nwparker in https://github.com/stablyai/orca/pull/33458 * refactor(native-chat): group tool attribution modules by @nwparker in https://github.com/stablyai/orca/pull/33482 * refactor: group native chat edit and patch modules by @nwparker in https://github.com/stablyai/orca/pull/33497 * refactor: organize native chat tool content and presentation by @nwparker in https://github.com/stablyai/orca/pull/33515 * refactor: group PTY bindings and lifecycle suites by @nwparker in https://github.com/stablyai/orca/pull/33566 * refactor: group renderer session launch lifecycle by @nwparker in https://github.com/stablyai/orca/pull/33586 * refactor: group local filesystem guards and search owners by @nwparker in https://github.com/stablyai/orca/pull/33624 * refactor(mobile): group native chat send and stream state by @nwparker in https://github.com/stablyai/orca/pull/33640 * test(connectors): drive a connector conversation end to end on a headless serve host by @Jinwoo-H in https://github.com/stablyai/orca/pull/33613 * Stabilize notification dispatch tests with fixed timestamps by @AmethystLiang in https://github.com/stablyai/orca/pull/33698 * test(git-bash): give ssh-agent a load-tolerant bound in the Windows pane proof by @nwparker in https://github.com/stablyai/orca/pull/33287 * test(windows): read the grandchild pid from accumulated pane output by @nwparker in https://github.com/stablyai/orca/pull/33320 * test(browser): route input into the baseline sandboxed frame before clicking it by @nwparker in https://github.com/stablyai/orca/pull/33330 * test(mobile-web): act on the HTML preview frame only once input reaches it by @nwparker in https://github.com/stablyai/orca/pull/33336 * test(windows): warm WMI before timing the NSIS capability probe by @nwparker in https://github.com/stablyai/orca/pull/33340 * test(mobile-web): click the drawer chip once it is actionable, and honour the wait timeouts by @nwparker in https://github.com/stablyai/orca/pull/33342 * test(terminal): load the alt-screen chain’s host graph at collection by @nwparker in https://github.com/stablyai/orca/pull/33343 * test: run released-phone compatibility checks under Node by @nwparker in https://github.com/stablyai/orca/pull/33480 * test(runtime): compare existing-folder imports against the resolved temp path by @nwparker in https://github.com/stablyai/orca/pull/33540 * test: honor startup paste timer delays and cancellation by @nwparker in https://github.com/stablyai/orca/pull/33550 * test: remove duplicate checks and repair weak fixtures by @OrcaWin in https://github.com/stablyai/orca/pull/33353 * test: remove redundant cases and strengthen regression assertions by @OrcaWin in https://github.com/stablyai/orca/pull/33356 * test: remove 13 low-value checks and strengthen runtime coverage by @OrcaWin in https://github.com/stablyai/orca/pull/33373 * test: remove redundant checks and repair audit coverage gaps by @OrcaWin in https://github.com/stablyai/orca/pull/33385 * test: remove redundant checks and repair misleading fixtures by @OrcaWin in https://github.com/stablyai/orca/pull/33394 * test: remove self-validating checks and repair behavioral coverage by @OrcaWin in https://github.com/stablyai/orca/pull/33412 * Remove low-value test inventories and repair behavioral assertions by @OrcaWin in https://github.com/stablyai/orca/pull/33437 * test: remove seven low-value cases and repair behavioral assertions by @OrcaWin in https://github.com/stablyai/orca/pull/33453 * test: remove schema inventory and strengthen boundary coverage by @OrcaWin in https://github.com/stablyai/orca/pull/33476 * test: exercise boundary assertions and repair audit fixtures by @OrcaWin in https://github.com/stablyai/orca/pull/33483 * test: verify retry identities and complete boundary outputs by @OrcaWin in https://github.com/stablyai/orca/pull/33491 * test: remove redundant checks and strengthen runtime assertions by @OrcaWin in https://github.com/stablyai/orca/pull/33498 * test: remove duplicate diff cases and repair audit fixtures by @OrcaWin in https://github.com/stablyai/orca/pull/33503 * test: remove mock-only cases and repair audit fixtures by @OrcaWin in https://github.com/stablyai/orca/pull/33509 * test: repair CLI and mobile fixtures and remove duplicate checks by @OrcaWin in https://github.com/stablyai/orca/pull/33518 * test: strengthen CLI and mobile behavior checks by @OrcaWin in https://github.com/stablyai/orca/pull/33526 * test: strengthen CLI errors and mobile lifecycle checks by @OrcaWin in https://github.com/stablyai/orca/pull/33529 * test: strengthen transport lifecycle and benchmark checks by @OrcaWin in https://github.com/stablyai/orca/pull/33538 * test: strengthen transport recovery and benchmark assertions by @OrcaWin in https://github.com/stablyai/orca/pull/33547 * test: tighten transport, recovery, and diagnostic checks by @OrcaWin in https://github.com/stablyai/orca/pull/33559 * test: strengthen relay, persistence, and benchmark checks by @OrcaWin in https://github.com/stablyai/orca/pull/33563 * test: tighten relay assertions and repair benchmark checks by @OrcaWin in https://github.com/stablyai/orca/pull/33573 * test: remove redundant checks and tighten runtime query coverage by @OrcaWin in https://github.com/stablyai/orca/pull/33583 * test: remove redundant metadata checks and strengthen runtime coverage by @OrcaWin in https://github.com/stablyai/orca/pull/33611 * test: remove redundant renderer and shared checks and strengthen assertions by @OrcaWin in https://github.com/stablyai/orca/pull/33625 * test: strengthen Git and Linear assertions and remove redundant checks by @OrcaWin in https://github.com/stablyai/orca/pull/33633 * test: strengthen watcher and mobile async assertions by @OrcaWin in https://github.com/stablyai/orca/pull/33641 * test: strengthen filesystem and settings coverage; remove redundant cases by @OrcaWin in https://github.com/stablyai/orca/pull/33652

New Contributors

Full Changelog: https://github.com/stablyai/orca/compare/v1.4.240...v1.4.241

Source: README.md, updated 2026-10-26