Meta just published the tech report for Muse Glimmer, its 30B open model built for agents that run on your own machine.
Apache 2.0, reads images, no API key.
Here's how to get it running locally without losing an hour 馃У
Apache 2.0, reads images, no API key.
Here's how to get it running locally without losing an hour 馃У
1
0
0
34
5/ Turn on the drafter for speed
The optional DFlash drafter (~1.6 GB) guesses 16 tokens at a time and the main model checks them. Same output, faster.
Meta's numbers: RTX 5090 went from 74.9 to 233.4 tok/s, M5 Max from 26.6 to 50.2.
The optional DFlash drafter (~1.6 GB) guesses 16 tokens at a time and the main model checks them. Same output, faster.
Meta's numbers: RTX 5090 went from 74.9 to 233.4 tok/s, M5 Max from 26.6 to 50.2.
1
0
1
19
6/ Two gotchas
Thinking can't be turned off. You set how much with a system line like "Reasoning strength: high" (low, medium, high, xhigh). Meta suggests high or xhigh for coding.
On llama-server, -c is split across slots. Run out and you get no answer, not an error.
Thinking can't be turned off. You set how much with a system line like "Reasoning strength: high" (low, medium, high, xhigh). Meta suggests high or xhigh for coding.
On llama-server, -c is split across slots. Run out and you get no answer, not an error.
0
0
0
5
5/ Keep it running in the background
On Linux or macOS, run t3 service install, then t3 service status to check it.
t3 update asks before restarting, because a restart cuts off running agent turns. On a Mac the service stops when you log out, so stay logged in.
On Linux or macOS, run t3 service install, then t3 service status to check it.
t3 update asks before restarting, because a restart cuts off running agent turns. On a Mac the service stops when you log out, so stay logged in.
1
0
0
9
6/ Go in with the right expectations
The README says it plainly: very early, expect bugs. Big feature contributions aren't being accepted yet.
Try it on a side project first, with approvals on, before you point it at code that matters.
The README says it plainly: very early, expect bugs. Big feature contributions aren't being accepted yet.
Try it on a side project first, with approvals on, before you point it at code that matters.
0
0
0
7
3/ Change the default permissions
New threads start on Full access, so commands and file edits run with no prompts.
Fix it under Settings, General, New threads, Permissions. Pick Supervised, or Auto, where the provider's review approves routine stuff and asks about the rest.
New threads start on Full access, so commands and file edits run with no prompts.
Fix it under Settings, General, New threads, Permissions. Pick Supervised, or Auto, where the provider's review approves routine stuff and asks about the rest.
1
0
0
7
4/ Code from your phone
The agents keep running on your machine. Your phone is the remote.
Simplest path is T3 Connect: on the host, open Settings, Connections, sign in and turn it on. On a CLI host, run t3 connect. The host has to stay on and reachable.
The agents keep running on your machine. Your phone is the remote.
Simplest path is T3 Connect: on the host, open Settings, Connections, sign in and turn it on. On a CLI host, run t3 connect. The host has to stay on and reachable.
1
0
0
12
T3 Code is everywhere tonight. It's a free, open-source app that runs the coding agents already on your machine: Claude Code, Codex, Cursor, Grok Build, OpenCode, Antigravity.
Bring your own subscriptions, steer from desktop, web or phone.
How to set it up without regrets 馃У
Bring your own subscriptions, steer from desktop, web or phone.
How to set it up without regrets 馃У
1
0
0
50
2/ Install it
Mac: brew install --cask t3-code
Windows: winget install T3Tools.T3Code
Or grab the desktop app from the project's GitHub releases page. If you go the CLI route, run t3 to start the server and open the local web app.
Mac: brew install --cask t3-code
Windows: winget install T3Tools.T3Code
Or grab the desktop app from the project's GitHub releases page. If you go the CLI route, run t3 to start the server and open the local web app.
1
0
1
19
1/ Log in to your agents first
T3 Code doesn't come with its own model. It drives the CLIs you already pay for.
Set up at least one before you open it: codex login for Codex, claude auth login for Claude Code, agent login for Cursor's CLI.
T3 Code doesn't come with its own model. It drives the CLIs you already pay for.
Set up at least one before you open it: codex login for Codex, claude auth login for Claude Code, agent login for Cursor's CLI.
1
0
0
24
4/ Say what you didn't test
The same doc tells AI to state plainly what couldn't be built, tested, or reproduced.
Put that line in your message too. "Didn't test on Windows" can save a reviewer an hour.
The same doc tells AI to state plainly what couldn't be built, tested, or reproduced.
Put that line in your message too. "Didn't test on Windows" can save a reviewer an hour.
1
0
0
5
5/ Tag the AI, sign as yourself
The kernel asks for an "Assisted-by:" line naming the model and any analysis tools. An AI agent must never add Signed-off-by. Only a human can sign off.
You reviewed it, so you own it.
The kernel asks for an "Assisted-by:" line naming the model and any analysis tools. An AI agent must never add Signed-off-by. Only a human can sign off.
You reviewed it, so you own it.
1
0
1
4
6/ Cut the filler
Delete "This commit...", "improves," "enhances," and "various fixes." If the subject could describe any commit, rewrite it.
Quick test: could a teammate find this commit with git log --grep a year from now?
Delete "This commit...", "improves," "enhances," and "various fixes." If the subject could describe any commit, rewrite it.
Quick test: could a teammate find this commit with git log --grep a year from now?
0
0
0
3
3/ Borrow the Linux kernel's template
Its guidance for AI coding assistants asks for a detailed message that covers the problem, the solution, and a Fixes tag pointing to the commit that caused the bug.
That works in any repo, not just the kernel.
Its guidance for AI coding assistants asks for a detailed message that covers the problem, the solution, and a Fixes tag pointing to the commit that caused the bug.
That works in any repo, not just the kernel.
1
0
0
6
A dev's post took off today: "AI assisted coding is okay, but wtf is your commit message AI generated?"
Fair point. A git log full of "Update logic and improve handling" helps nobody six months from now.
How to use AI on commit messages and keep the part that matters 馃У
Fair point. A git log full of "Update logic and improve handling" helps nobody six months from now.
How to use AI on commit messages and keep the part that matters 馃У
1
0
0
11
2/ Give it context, not just the diff
Paste the issue, the error, or your notes next to the output of git diff --staged.
Then ask: "Write a commit message. Short subject. Body covers the problem, the fix, and anything risky."
Paste the issue, the error, or your notes next to the output of git diff --staged.
Then ask: "Write a commit message. Short subject. Body covers the problem, the fix, and anything risky."
1
0
0
1