New
Everything new in AI as it appears: launches announced by the labs, brand-new models, datasets, papers and open-source tools. Checked every minute or two. Launches feed · Trending · News
Live now
- Connecting…
Today
- Tool@ai-sdk/google
The **[Google provider](https://ai-sdk.dev/providers/ai-sdk-providers/google)** for the [AI SDK](https://ai-sdk.dev/docs) contains language model support for the [Google
- Tool
- Tool@ai-sdk/moonshotai
The **[Moonshot AI provider](https://ai-sdk.dev/providers/ai-sdk-providers/moonshotai)** for the [AI SDK](https://ai-sdk.dev/docs) contains language model support for the
- Tool@ai-sdk/mistral
The **[Mistral provider](https://ai-sdk.dev/providers/ai-sdk-providers/mistral)** for the [AI SDK](https://ai-sdk.dev/docs) contains language model, embedding model, and
- Tool@ai-sdk/code-mode
QuickJS-backed code mode tool for AI SDK
- Tool@ai-sdk/huggingface
The **[Hugging Face Inference Providers](https://huggingface.co/docs/inference-providers/index)** for the [Vercel AI SDK](https://ai-sdk.dev/docs) contains language model
- Tool@ai-sdk/harness-opencode
The OpenCode harness connects `HarnessAgent` to OpenCode through a sandboxed bridge.
- Tool@ai-sdk/azure
The **[Azure provider](https://ai-sdk.dev/providers/ai-sdk-providers/azure)** for the [AI SDK](https://ai-sdk.dev/docs) contains language model support for the Azure Open
- Tool@ai-sdk/harness-codex
`HarnessV1` adapter backed by [`@openai/codex-sdk`](https://www.npmjs.com/package/@openai/codex-sdk), which drives the `codex` CLI. The adapter ships a bridge process tha
- Tool@ai-sdk/groq
The **[Groq provider](https://ai-sdk.dev/providers/ai-sdk-providers/groq)** for the [AI SDK](https://ai-sdk.dev/docs) contains language model support for the Groq chat an
- Tool@ai-sdk/harness
_This package is **experimental**._
- Tool@ai-sdk/harness-claude-code
`HarnessV1` adapter backed by [`@anthropic-ai/claude-agent-sdk`](https://www.npmjs.com/package/@anthropic-ai/claude-agent-sdk), which drives the `claude` CLI. The adapter
- Tool@ai-sdk/harness-acp
`HarnessV1` adapter backed by an NPM-installed [Agent Client Protocol](https://agentclientprotocol.com/) version 1 implementation. The adapter ships a bridge process that
- Tool@ai-sdk/cerebras
The **Cerebras provider** for the [AI SDK](https://ai-sdk.dev/docs) contains language model support for [Cerebras](https://cerebras.ai), offering high-speed AI model infe
- Toolpi-web-ui
Just open your browser to get all your work done — browser cockpit for AI coding agents (chat, code, files, terminal, Git in one tab)
- Tool@sleds/cli
SLEDS CLI — the cost and limits of AI agent work, decided at the wall. Point Claude Code at a SLEDS proxy; estimate before you run; hold, switch model, or run overnight w
- Toolmomentic-mobile
The official CLI for Momentic Mobile. Build, run, and manage end-to-end tests for iOS and Android through natural language prompts.
- Tool@devvibex/aiflow
First-party LLM gateway SDK for apps generated on the VibeX platform — chat, streaming, image, video (including stateful multi-turn video extension with Gemini Omni), mus
- Tool@social-data/social-neuron-opinion-query
Social Neuron opinion-query 能力 CLI —— 一能力一命令(social-neuron-opinion-query),薄客户端直达 /opinion-query/{op}。
- Tool@clawdbot/lobster
Workflow runtime for AI agents - deterministic pipelines with approval gates
- Toolqa
The CLI for Mo, Momentic's autonomous QA agent.
- Toolmomentic
The official CLI for Momentic. Build, run, and manage end-to-end tests through natural language prompts.
- Tool@azure/mcp
Azure MCP Server - Model Context Protocol implementation for Azure
- Tool@azure/mcp-linux-x64
Azure MCP Server - Model Context Protocol implementation for Azure, for linux on x64
- Tool@azure/mcp-win32-x64
Azure MCP Server - Model Context Protocol implementation for Azure, for win32 on x64
- Tool@azure/mcp-darwin-arm64
Azure MCP Server - Model Context Protocol implementation for Azure, for darwin on arm64
- Tool@layer_ai/sdk
Official JavaScript and TypeScript SDK for Layer — AI asset creation for game studios.
- Tool@thomasfosterau/effect-evals
An Effect-native LLM evaluation toolkit — goldens, scorers, thresholds, metrics (G-Eval, RAG, retrieval ranking: nDCG / recall / MRR / precision@k), a caching runner, and
- Toolproactive-gate
proactive-gate is a library that decides whether an AI agent may message a user right now (quiet hours, daily limit, consent, snooze) and records the reason when it may n
- Toolagentfootprint
The explainable agent framework — backtrack a wrong answer to the exact context that caused it (evidence, not guesses). Built on footprintjs.
- Tool@chai-ui/core
Chai UI headless core — the Request/Run/Result data model and pluggable generation engines (Fal.ai, OpenRouter, Vercel AI Gateway, mock). No framework, no rendering.
- Tool@chai-ui/vue
Chai UI Vue components — smart, drop-in UI for AI apps: Composer, Media Analyzer, Result Card, Edit Card, and the useComposer composable.
- Tool@chai-ui/react
Chai UI React components — smart, drop-in UI for AI apps: Composer, Media Analyzer, Result Card, and the useComposer hook.
- Tool@chai-ui/tokens
Chai UI design tokens — single source of truth, built to CSS variables, TypeScript, and a Figma Tokens/Variables-plugin JSON.
- Tool@dreamworkhq/mcp
Job search, application materials, and application tracking for AI assistants.
- Tool@thomasfosterau/effect-ai-voyage
Effect AI provider for Voyage AI — EmbeddingModel and RerankerModel implementations over Voyage's embeddings and rerank APIs.
- Toolopencode-mem
OpenCode plugin that gives coding agents persistent memory using local Turso/libSQL vector search
- Tool@kenz1117/dsh-ui-usage-billing
Usage billing dashboard for DeepSeek Harness: sidebar cost metrics plus a full dashboard modal, priced from a current multi-provider catalog with real usage aggregated fr
- Tool@miller-tech/uap
Autonomous AI agent memory system with CLAUDE.md protocol enforcement
- Tool@foldkit/devtools-mcp
MCP server exposing Foldkit DevTools to AI agents (Claude Code, Cursor, etc.)
Headlines and links come from the labs' own blogs, Hacker News and the GDELT news index, and from the sources our index already follows. We show a short excerpt and always link to the original. A launch reported by a news site rather than the lab is marked “Reported” until the lab itself announces it.