Our open source Agent Chat UI now offers standardized support for file uploads! https://t.co/PTNeLInRrn
— LangChain (@LangChain) 20 mai 2025
Our open source Agent Chat UI now offers standardized support for file uploads!

By
–
Our open source Agent Chat UI now offers standardized support for file uploads! https://t.co/PTNeLInRrn
— LangChain (@LangChain) 20 mai 2025
Our open source Agent Chat UI now offers standardized support for file uploads!
By
–
Agent Chat UI: https://
agentchat.vercel.app GitHub Repo:

By
–
New Keras release is out — includes saved model file sharding for very large models, the Muon optimizers, new ops and many performance improvements
By
–
Weights, code, and paper @HuggingFace
: https://
huggingface.co/stabilityai/sv
4d2.0
… @Github
: https://
github.com/Stability-AI/g
enerative-models
… @arXiv
: https://
arxiv.org/abs/2503.16396 (4/4)

By
–
Turn internet into an agent-friendly environment just using natural language! Notte is the open source framework that allows you to develop, deploy, and scale your own agents with a single API. 100% Open Source
By
–
Optionally, you can add a http://
PROMPT.md file to override the default Agent prompt. Then just point tiny-agents to your local folder: npx @huggingface/tiny-agents run ./my-agent Voilà!
By
–
Link to the code: https://
github.com/huggingface/hu
ggingface.js/tree/main/packages/tiny-agents
… PS: this is built on top of @huggingface partners-powered Inference Providers, and our MCP client

By
–
Today I'm releasing Tiny Agents into its own NPM package It is a squad of lightweight composable agents, built on Hugging Face's Inference Client and MCP stack. Here's a tiny tutorial to get started.
By
–
Creating your own agent is as easy as creating a folder containing an agent.json file: mkdir my-agent
touch my-agent/agent.json Inside the JSON file, set a model and a provider (or a local endpoint), and a list of MCP servers: {
"model": "Qwen/Qwen2.5-72B-Instruct",

By
–
Is compression really linear with code intelligence?
A research team from China dives deep into this question.
They evaluated a wide range of open-source Code LLMs across multi-language, multi-task benchmarks — and introduced Format Annealing for fair comparison. Key