CoreWeave Developers

Official updates for developers building on @CoreWeave. CoreWeave Forge, Weights & Biases Models, inference, sandboxes, and evals.

[CoreWeave Developers @coreweavedevs]

Introducing W&B Skills for coding agents!

Watch us install the skill, point it at a live fine-tuning project with thousands of traces and RL training runs, then query it all from the terminal in seconds.

Also works with @weave_wb to pull in your agent traces and spot failures.

[Lorenzo Porras @TheCodingSoup]

Swarms are everywhere, so I got curious how the agents actually talk to each other.

Gave each agent its own @CoreWeave Sandbox, traced every message in Agent Lens, and added a shared memory. They wrote 12 notes. Nobody read a single one.

Also really enjoying gamifying all my demos lately 🎮

More to come on this fun one.

[CoreWeave Developers @coreweavedevs]

@TheCodingSoup wanted to see how AI agents talk to each other when they work as a team.

So he gave a swarm a shared memory and let them do their thing inside of a @CoreWeave Sandbox. They wrote 12 notes. Not one got read by another agent. (Every agent was traced in Agent Lens, so we could check.)

The next run changed the timing: leads shared plans before starting, and each new agent got the relevant notes with its task. No duplicated experiments that time.

Try out Agent Lens https://t.co/hkBJpJRmwc

[CoreWeave Developers @coreweavedevs]

Karpathy got annoyed with YouTube recs and built HackerTube in a weekend. Yours was probably manually lining up logged images to check a single detection.

New in Weights & Biases Models: zoom or pan one image and the rest follow at the same zoom level and position. Bounding boxes, masks, and class labels move with their images.

It's on by default, so the cropped screenshots can retire. https://t.co/y6yFm2tZaO

[Geoff Pado @cocoatype]

@wandb Turn it over into “tent mode” and you can have a fun little auto-updating metric dashboard sitting on your desk.

[Reflection @reflection_ai]

Lorenzo Porras (@thecodingsoup) built an escape room where Beam repairs a planetary machine, routes light through mirrors, and plans a robot's escape across bridges and doors. Beam receives the rules and state through tools, then writes and runs Python in @CoreWeave Sandbox to solve the puzzles. A 3D room shows the run as it happens.

[CoreWeave Developers @coreweavedevs]

@reflection_ai @TheCodingSoup @CoreWeave @TheCodingSoup

[CoreWeave Developers @coreweavedevs]

@Vanzllll We got you! 🫡

[CoreWeave Developers @coreweavedevs]

wandb 0.31.0 is out now, and import wandb and the CLI start about twice as fast.

That's for builders like @Vanzllll, whose DiffusionART team "found some surprising answers" after nearly 800 wandb runs.

Now find yours.

It also grows W&B LEET, the terminal UI for exploring W&B runs. wandb leet inspect --summary reads a run's local .wandb file, and --json prints that as JSON for a coding agent.

pip install -U wandb

[CoreWeave Developers @coreweavedevs]

iPhone Duo is out today, and Weights & Biases Models already supports 2-fold validation 📲

Folded, the cover screen tells you the run is still alive.

Unfold it, and your runs sit next to your panels on the big inner screen.

Hide a run, and every chart redraws right away, making the one that spiked much easier to find.

The app is free and ready for your Duo on day one. https://t.co/VaG0BQLhe7

[CoreWeave @CoreWeave]

Richard Ahlfeld walked @altryne through how he explains physical AI to LLM engineers.

The reasoning a good chatbot does stays the same.

Our CoreWeave Physical AI team just released Launchables.

Tested Templates for robotics and simulation tools, each run with one command. https://t.co/SUnFk9EZ9p

[CoreWeave Developers @coreweavedevs]

Your policy nails it in sim and falls over on the real floor. Bring that one to Rainey Street.

Join us and @NVIDIARobotics for a night at CoRL in Austin, Nov 10, 6 to 9 PM CT.

Our team will give you a straight technical read on what you're building.

https://t.co/LlCX53Xfdq

[CoreWeave @CoreWeave]

Five weeks ago, Jev was a secret on our San Francisco floor 🥷

@typesafeai was still in stealth when it sponsored CoreWeave Hacks on September 12 and 13. The team brought Jev before it launched and stayed all weekend helping over 60 teams build on it.

Two days later, @CompleteSkeptic announced Jev to his 1,500 followers. That post is now past 40 million views, and he's at 156K followers.

It hasn't slowed down since.

He told the WSJ that Jev had now crossed a trillion tokens a day and was in use by about a quarter of the Fortune 500.

Workday's new Workday Home is powered by it, and Vercel says nothing was ever adopted faster on its AI Gateway.

Proud that they unveiled it with our builders first. This one's going in the CoreWeave Hacks history books.

[CoreWeave Developers @coreweavedevs]

We knew Beam back when it was just a loss curve 🍼

[CoreWeave @CoreWeave]

🔦 Beam, Reflection's first model with open weights, is live today on CoreWeave Serverless Inference.

Call it through an API and leave the GPUs to us.

We are also proud to share that @reflection_ai used @wandb in CoreWeave Forge during Beam's training.

Developers in the private preview have been calling it fast and accurate, and they like that it doesn't ramble.

See what @TheCodingSoup on our team built with private preview access below.

Get started with Beam here: https://t.co/tyRVazUtUm

[CoreWeave Developers @coreweavedevs]

New model dropping soon on @CoreWeave Serverless Inference 👀

CoreWeave Agent Lens (@weave_wb) tracing built in.

[Lorenzo Porras @TheCodingSoup]

During @reflection_ai’s private preview of Beam, I built a haunted escape room to put its coding and reasoning to the test.

I created a game called “Code Your Way Out”, and it's one room with three linked puzzles. Beam has to repair a brass orrery by lining its rings up with a constellation, route the light, then program its escape through a sealed hatch.

Beam never sees the room. Each turn, it gets text observations and writes Python; the code runs in a @CoreWeave Sandbox, and the game engine validates every action before the explorer robot moves.

Beam was out in about 30 seconds. Super bullish on this model and really excited to see builders get this model in their hands.

[CoreWeave @CoreWeave]

The @NVIDIAAI team shipped Nemotron 3.5 Super VL today, and you can call it now on CoreWeave Serverless Inference.

Send it text and images, build with supported reasoning, tool calling and structured outputs, and trace every call in CoreWeave Agent Lens.

It's a multimodal mixture-of-experts model with 12B active parameters, served in NVFP4.

Pricing per 1M tokens is $0.25 input, $1.50 output, and $0.05 cached input.

Try it here: https://t.co/MUfcALKOnu

[Aaron @aaronbatilo]

@CoreWeave Agent Lens already supports OTel, natively. All you have to do is configure your endpoint and we'll slurp all your Claude traces into CoreWeave Forge's purpose built UI. You can do this literally right meow.

https://t.co/19D3f4xoIl

[CoreWeave Developers @coreweavedevs]

@ted_engineer @chaotictransfem @CoreWeave @OpenPipeAI @weave_wb 😂

[CoreWeave Developers @coreweavedevs]

@chaotictransfem @CoreWeave @OpenPipeAI @weave_wb Whose to say we can’t still use it? wandb forever.