Run any command on a remote GPU from your terminal
Boulder, Colorado
Joined April 2025
- Tweets104
- Following106
- Followers28
- Likes679
GPU CLI retweeted
Really enjoying working with the new #attn changes that allow me to share thoughts on html files.
They're agent agnostic and allows me to quickly provide artefact feedback.
Struggling to build Swift UIs that don't look like slop?
We built AnnotKit to help. It's a small tool that annotates your macOS/iOS dev build in place. Just point at a view, type a note, and your agent knows exactly what to change and where.
annotkit.gpu-cli.sh
GPU CLI retweeted
Really enjoying using the new `clean-comments` skill we made. It's helped reduce a lot of noise in my recent PRs
Tired of the overly wordy responses your agents throw at you? We just published a new skill called `less` that will help you get to the heart of your issues quicker.
To use it, just install it and specify the number of sentences you'd like to compress a response to and whether you want to set this as the response length moving forwards too.
You can find this skill and many other gems in our dedicated repo below.
github.com/gpu-cli/skills
GPU CLI retweeted
Windows development is actually fun
- @unity 's Parsec + Windows Terminal + Claude Code (or Codex)
- No RAM bloat from WSL running in the background.
Mostly been working on inference for windows.
GPU CLI `0.32.0` has native windows support so bursting into the cloud works nicely too
Step 3: If you're a team, check the gates.
Even above break-even, owning only wins for more powerful hardware if you also own the datacenter - or if latency, regulation, or supply scarcity force your hand.
Own the building and the H100 break-even drops to ~39–45%.
Run the steps and for most teams the answer is the same, rent, and ideally on a committed contract if usage is steady. Buy only when you clear the steps above.
Learn more here: gpu-cli.sh/blog/should-i-ren…
Searching for a faster way to find the best deal to rent that GPU you need?
gpu-cli.sh/rent just shipped 🚢
Now you can now compare prices from @runpod, @vast_ai, @ThunderCompute & more in one place.
GPU CLI retweeted
What was used to rewrite Bun into Rust apparently
GPU CLI retweeted
I wanted faster builds so I explored how to conditionally build our rust monorepo based on the cargo dependency graph - packaged it up as a github action for reuse across projects github.com/lightsofapollo/ca…
GPU CLI retweeted
To check if your Google Workspace has been compromised by the same tool that compromised Vercel:
1. Go to admin.google.com/ac/owl/list…
- This is Google Admin Console > Security > Access and Data Control > API Controls > Manage app access > Accessed Apps
2. Filter by ID = 110671459871-30f1spbu0hptbs6…
- This is the ID of the compromised OAuth app
If you see an app after filtering, you have potentially been compromised
GPU CLI retweeted
We're proud to share that @NVIDIA submitted the first-ever MLPerf Vision Language Model (VLM) performance benchmark using vLLM. This achievement showcases the strength of our ongoing collaboration with NVIDIA Engineering.
Check out their MLPerf blog and watch our On Demand Talk at GTC to learn more about how we are delivering the best performance on NVIDIA hardware.
🔗 Blog: developer.nvidia.com/blog/nv…
🔗 Talk: nvidia.com/en-us/on-demand/s…
Open-source software never stops. It only accelerates.
Dynamo, @sgl_project, TensorRT LLM, and @vllm_project are constantly optimized by a vast ecosystem of developers building on top of the NVIDIA platform.
The result: your token output keeps improving and token cost keeps decreasing on the same hardware resources while your developer velocity stays at its peak.
Build on the foundation continuously optimized by the world’s best developers. ⚡
🔗 nvda.ws/3OsTQL0