Skip to content

Releases: coder/mux

v0.12.2

13 Jan 01:11

Choose a tag to compare

This release is a hotfix to restore terminal input functionality (see #1586)

v0.12.1

12 Jan 16:31

Choose a tag to compare

Features

  • Login with Mux Gateway is now live!
    If you have a coupon code you can still redeem them at https://gateway.mux.coder.com/
    When launching mux you can login to the gateway to automatically use the vouchers.
image

You can also join our discord to talk mux and request some tokens for testing!

Fixes

  • improvements to review pane - file status, CRLF handling and inline review comments
  • revamp of the provider menu for easier management of custom models
image

Full Changelog: v0.12.0...v0.12.1

v0.12.0

09 Jan 15:46

Choose a tag to compare

🏙️ Highlights

We now support integrated terminals in the right sidebar, with drag-and-drop tiling and resizing:

image

Terminals persist on the backend, so mux server users can leave a tab, come back later, and resume the session exactly where they left off.


New runtime: Docker!

Screenshot 2026-01-09 at 9 41 01 AM

mux run has been given a makeover to make it useful for automation and CI. See:

Screenshot 2026-01-09 at 9 39 54 AM

🛠️ Fixes

  • Unified handling of environment variables (docs)
  • Better default $PATH on macOS using common shell profiles (49176ed39)

v0.11.1

05 Jan 02:11

Choose a tag to compare

Primarily a hotfix for 5dd8529 — fixing a rare bug where Anthropic reasoning would be effectively disabled sometimes on long-running chats.

Also, sneaks in e4e41f7 which embeds the Mux docs as a built-in skill. So, users can ask Mux about itself for support.

v0.11.0

04 Jan 20:48

Choose a tag to compare

🌃 Highlights

Project now support customizable sections for organization:
image

You can share plans and other assistant outputs from Mux via the e2e encrypted mux.md:

image

Idle-compaction settings moved next to usage-based auto-compact slider:

image

When reaching a context exceeded error, we now give you a more useful retry operation if you have a model available with a higher context limit:
image

See also:

  • Fully customizable Agents (docs)
  • Mux can notify via new notify tool, just instruct via prompt (docs)
  • Tool hooks! (docs)
  • You can tag files with "@" in the chat which will include them in your context

🛠️ Fixes

  • Fixed bug where SSH workspace sub-agents had unexpected git state (b4125da)
  • Fixed thundering herd issue where enabling idle compaction could lead to dozens of old workspaces undergoing simultaneous compaction
    • New behavior: only one idle compaction runs at a time (f24db30)
  • Many performance improvements all around the app

v0.10.1

29 Dec 12:45

Choose a tag to compare

What’s Changed

🚀 Features

  • Add a Keybinds section to Settings (improves discoverability of shortcuts).
  • Add simulateFileModifyingToolEnd for UI testing.
  • Add model-only TODO reminder notifications.
  • Add VS Code secondary sidebar mux chat view (better VS Code integration).
  • Add PTC subagent support plus compatibility improvements.

⚡ Performance

  • Prevent sidebar re-renders during streaming (reduces UI churn / improves responsiveness).

🐛 Fixes

  • Bug fixes around PTC, UI, etc.

v0.10.0 🎄

27 Dec 02:40

Choose a tag to compare

This release includes a wide range of features and bug fixes spanning the entire product.

✨ Highlights

We now have fledgling support for sub-agents. When you enter Plan Mode, the user-facing agent may spin up any number of sub-agents for exploration and research tasks.

Workspaces are now Archived instead of deleted by default. Users can easily revive archived workspaces with all of their state intact:

Screenshot 2025-12-26 at 7 35 34 PM

Archived workspaces continue to consume the same storage as live workspaces, but no longer clutter the UI or run background processes. This also makes it easier to track longer-term costs.


We shipped several improvements to the Code Review experience:

Screenshot 2025-12-26 at 7 39 58 PM
  • The review panel now refreshes automatically — edits stream in real time
    • Edits are sorted by first-seen timestamp, so the most relevant hunks appear at the top
  • You can load more of a file to see surrounding context
  • Review controls are significantly more compact

You can now more easily see context usage in the chat input area:

image

This replaces the infamous vertical usage bar that resembled a scrollbar.

🫡 Honorable Mentions

  • Many improvements to compaction, including better preservation of plan context, recent edits, and TODO lists
  • Support for agent skills (docs)

🛠️ Fixes

  • Significant performance improvements across the app
  • Reduced layout flashing and smoother transitions
  • Improved compaction robustness, with better behavior around interruptions and edits

v0.9.7

14 Dec 22:27

Choose a tag to compare

This release primarily contains polish and performance improvements.

Screenshots

Revised tabs and diffs:
image

Features

  • Added branch selector to Workspace title bar
  • Major performance improvements for long chat histories
  • Revised thinking level to persist by model
  • Revised Model Switcher to support use case where you have a small number of "go-to" models that you frequently switch between

v0.9.6

12 Dec 11:36

Choose a tag to compare

Features

  • Post-compaction context preservation for better continuity.
  • Cumulative cost tracking for the entire chat
  • Shows line additions/deletions in the git status indicator, with a toggle to enable/disable.
  • OpenAI service tier improvements
    • Defaults the OpenAI service_tier to Auto
    • Adds a service tier dropdown in the UI

Fixes

  • Model display with mux-gateway: Keeps model display formatting clean and consistent when using mux-gateway.
  • New chat focus shortcut: Adds Ctrl+I to focus the input on the new chat page.
  • Gateway toggles + model caching defaults: Fixes gateway toggle behavior and adjusts model LRU defaults.
  • Editor deep link fallback: Falls back to editor deep links when the CLI isn’t available.

v0.9.5

11 Dec 19:16

Choose a tag to compare

Hot-release for GPT 5.2 models.