Personalize your Framework Desktop with high quality customized tiles!

Utah, USA
Joined August 2025
Forgework retweeted
We did it! Framework Laptop 13 Pro is on TIME's Best Inventions of 2026. This is our third product to list, after the original Framework Laptop in 2021 and Framework Laptop 16 in 2023.
51
90
17
2,064
29,947
Forgework retweeted
Our current stack for Halo Strx local LLM serving. Runtime -Gufo Model - unsloth’s qwen3.8 flash next q4_k_xl Harness - Hermes 🪽 1300tok/s encode and 40tok/s decode on a model that beats deepseek v4.1 flash. We have arrived.
4
1
7
5,730
Forgework retweeted
Replying to @sivaemon
Hello, electrical expert here! This is not cute or funny, cables only do this when extremely stressed!
8
12
2
649
8,231
Did you hear the news?! Imagine the shit we'll be able to do with shell now! Incredibly excited to get to work with @outfoxxedd to really shape the future of Omarchy Shell.
6
4
215
8,649
Forgework retweeted
My two faves
Hermes Agent is now built into Omarchy. what it means in practice: the desktop app installs from the AI menu like any other Omarchy app, or you make Hermes your default terminal agent, and it picks up your Omarchy theme, desktop app, TUI and CLI in the same colours. one machine, one agent, every surface. a very natural home for it.
1
1
2
127
Forgework retweeted
Framework tiles :) @Forgework_ @FrameworkPuter
1
4
59
Forgework retweeted
The new king of local models is out!
We promised open weights for Qwen3.8. Now, time to meet them! 🎉 ⚡ Qwen3.8-27B: - A native multimodal dense model. With just 27B parameters, it outperforms Qwen3.7-Plus overall and shines in real-world coding & office workflows. - 262K native context, easily extendable to 1M tokens via YaRN. - Built for builders. Highly efficient, high-quality, and licensed under Apache 2.0. 🚀 The open weights for Qwen3.8-2.4T-A95B (Max-level) have also been released recently. Whether you're shipping lightweight applications with Qwen3.8-27B locally or building agents with Qwen3.8-2.4T-A95B, they're yours now! Download, deploy, and build something we haven't imagined yet. 👀👇 - Hugging Face: huggingface.co/collections/Q… - ModelScope: modelscope.cn/collections/Qw…
40
55
1
1,281
63,277
Forgework retweeted
We keep throwing it out, and every morning it appears back on our desktop.
1
2
99
Forgework retweeted
Replying to @sudoingX
Concur in all respects!
2
1
2
214
Forgework retweeted
We sent a bunch of fun tiles over to @hwmedialabllc because we had to see what they looked like on the DockFrame!
LEGO studs on a USB-C dock. Why? So two DockFrames can stack for more slots, and so a clip-on frame can carry the same tiles you put on your @FrameworkPuter Desktop. Big thanks to @Forgework_ for the test set. hwlab.io/blog/lego-studs-usb… #dockframe #opensource #framework
1
3
11
1,432
Forgework retweeted
This is enough to make someone drool…
1
1
14
596
cant wait to sell b2b saas to aliens
25
36
8
286
12,707
Forgework retweeted
Replying to @AMDRyzen @lmstudio
You need some tiles from @Forgework_
1
2
120
Forgework retweeted
My "1000 robot" inference army is growing . . . @FrameworkPuter is awesome and special thanks to @Forgework_ for the wicked custom tiles!!! Now if only I could get a qty discount for PLEX (tiles already secured)!
1
4
256
Forgework retweeted
Okay I’ll do it. I’m taking the battery out. Prepare yourselves.
New Expansion Card idea @i2cjak
17
11
1
973
27,891
Forgework retweeted
New Expansion Card idea @i2cjak
5
52
1
3,330
74,316
Forgework retweeted
day 1 of designing more and more complicated framework *EXPANSION CARDS until they give me a free laptop Today, a Wi-Fi HaLow expansion card! This new(er) Wi-Fi standard uses sub-gig operating frequencies for superior range! LoRa but not as long range but 100x the throughput!
day 0 of designing more and more complicated framework modules until they give me a free laptop
73
67
4
2,801
172,571
Forgework retweeted
Time to switch things up on your Framework Desktop? We can help: forgework.shop :)
2
2
2
19
2,944
Forgework retweeted
LM Studio 0.4.12 is out now - @Alibaba_Qwen Qwen3.6 support! - Nicer PDF exports for chats - MCP servers with OAuth work on Windows - Better API performance for Qwen3.5
21
54
6
598
26,932
Forgework retweeted
how to speed up your framework desktop performance: 1. setup 1280GiB GTT 2. use MoEs (10B active or less is best) 3. Vulkan RADV 29% faster decode 4. the smaller the active params the faster it go
7
6
1
153
43,585