2026-10-21 AI 日报

DeepInfer: High-performance, large-scale inference framework for DeepSeek models

The cross-harness, self-improving memory layer for AI agents. Join our community: https://discord.com/invite/zdNChS2fBu

Turn any codebase, with its docs, SQL schemas, configs, and PDFs, into a queryable knowledge graph. A /graphify skill for Claude Code, Cursor, Codex, and…

Orca is the place for working with a fleet of agents (e.g. Grok Bot, Claude Code). Run any coding agent with your own subscription. Available on desktop,…

Makes your AI agent think like the laziest senior dev in the room. The best code is the code you never wrote.

Free, open-source AI Office suite: Docs, Sheets, Slides, PDF, Markdown and HTML editors with a built-in AI agent, plus a `genoffice` CLI and agent skill so…

Compress tool outputs, logs, files, and RAG chunks before they reach the LLM. 20% fewer tokens for coding agents, 60-95% fewer tokens for JSON, same answers.…

7.4 billion tokens per month. 34 free LLM providers. 635 free model endpoints. All behind one /v1 endpoint, plus any custom OpenAI-compatible endpoint. Smart…

The most powerful and modular diffusion model GUI, api and backend with a graph/nodes interface. The fastest local inference engine in the world.

"Vibe-Trading: Your Personal Trading Agent"

In this report, we introduce Sage-A1.5, a 35B Mixture-of-Experts agentic model that builds on A1 to advance long-horizon problem solving through adaptive…

World models enable embodied agents to anticipate action consequences, yet jointly modeling locomotion and manipulation across diverse embodiments remains…

Spatial reasoning asks a model to work out where objects are in 3D space, how they relate to one another, and how they move. Agentic VLMs have recently…

Game engines and coding agents can specify the layout, motion, and camera of a scene, yet turning such coarse structure into realistic video requires a…

Model--harness co-evolution has emerged as a prominent paradigm for recursive self-improving (RSI) agents, jointly evolving their execution scaffolds and…

Computer-use agents are limited by the interfaces humans give them. An agent may type a spreadsheet cell by cell when a single line of code would do, or print…

Moving from controllable videos to genuinely playable worlds requires persistent world state, executable game rules, and low-latency interaction. We outline a…

Looped language models can run more iterations at test time, but additional iterations need not help them solve longer problems. For sequential state…

Iterative sampling in diffusion models can lead to error accumulation due to exposure bias. The denoiser is trained on real images interpolated with noise,…

Flow-matching vision-language-action (VLA) models generate an action chunk by iteratively denoising Gaussian noise, but typically execute only a short prefix…

https://patchtheplanet.org/

Today, we’re launching a public preview of Mistral Large 4. Unofficially ML4, very officially: le Chonk. ML4 pushes the frontier of open-weight performance.…

Hey all, Kartik from Omnara (https://omnara.com) here. We're building an open-source API for running durable agents. You create an agent with an API call, and…

Anthropic is a Public Benefit Corporation (PBC), a designation that requires us to balance the financial interests of shareholders with the benefit to the…

This model looks really impressive. TL;DR -- Coding agent model. 40B params with 2.5B active. Resoundingly beats Qwen3.6-35B-A3B in TerminalBench 2.1 and…

DeepSeek recently shared some numbers showing that it processes around 155 trillion tokens per day and makes about $38 per GPU per hour. Using those numbers…

KittenML has just released code and weights of Kitten-TTS 2, which is a super-efficient frontier TTS model, and the world's first 1.58-bit TTS model . It runs…

Mixture-of-experts (MoE) models often exceed a single accelerator's memory capacity. Offloading experts to host memory creates a severe PCIe I/O bottleneck…

I held off as long as I could, but finally gave in and set up strata. I am pretty blown away. 3.8 27b at Q6 kv8 was my driver before and that works super…

I don't have a 3090 to confirm, but curious if the anyone in this community can back this up. Good win for 3090 and curious if that translates to other GPUs.…

Read relevant sections. Reuse known context. Check whole-file hashes when work resumes. I'm part of UPHASH. We tested Obsidian Agent Kit (MIT) with Codex,…

What forward deployed engineering means at OpenAI

See how OpenAI, PerturbAI, and Scripps Research used distributed AI agents to uncover promising research leads in data from millions of brain cells.

OpenAI and Google Cloud are bringing frontier models and Codex workflows to Gemini’s Model Garden, available through Google Cloud Marketplace.

We've released API version 2026-10-21. Migrating to API version 2026-10-21 shows how to upgrade.

We've released new versions of the client SDKs and the ant CLI that send API version 2026-10-21 by default: Python SDK 2.0.0, TypeScript SDK 1.0.0, C# SDK…

An ultra-efficient open-weight model for everyday work.

The Agents CLI Extension System enables developers to use frameworks like LangChain, LangGraph, or custom enterprise stacks while retaining a unified…

Voxtral 3: Mistral's voice agent SDK for real-time speech in 50+ languages — accurate transcription, emotional TTS, deep tool use. EU-hosted or self-deployed.