Step 5 Preview is now open weights. BF16 and FP8 weights are live on Hugging Face and ModelScope, together with our technical blog. • Sparse MoE: 600B total, 27B active per token • 1M-token context window • Native text and vision input • Built for long-horizon agentic work: coding, software engineering, professional knowledge work and finance Frontier intelligence, open weights. Build what's next. Hugging Face: ModelScope: GitHub:

FLUX 3 Image Klein is here. Open Weights with the quality of a large frontier model. Customizable and fast. 5.6B parameters and runs on a single consumer GPU. Native 4K, bounding boxes, and pixel-perfect editing.

Introducing Kylon: The first company harness for your humans, agents and tools. Whether you’re using Codex, Claude, Devin, Hermes, Gemini CLI, OpenClaw, Muse, Instinct, Grok Bot, Dots, Perplexity Computer, Cursor, Antigravity, Microsoft Autopilot, Amp, Droid, Kiro, Cline, OpenCode, Replit Agent, Lovable, v0, Gmail, Slack, Apollo, Supabase, Apify, Treg, Neon, Vercel, Databricks, Novita, RadixArk, GitHub, Seedance, Fish Audio, Plaud, OpenRouter, Gemini, Google Sheets, TikTok Shop, Outlook, Google Drive, Instagram, Airtable, Google Calendar, Asana, HubSpot, Corgi, SharePoint, Google Tasks, Notion, Google Workspace, Telegram, LINE, Zoho Mail, Meta Ads, Google Search Console, Fathom, ClickUp, Reddit, Granola, Google Docs, OneDrive, tl;dv, Amplitude, Dropbox, YouTube, Ahrefs, Xero, HighLevel, Google Analytics, Google Slides, Monday, QuickBooks, ElevenLabs, Linear, PostHog, LinkedIn, Runway, Zoho Bigin, Stripe, Exa, Cal_com, Box, Google Meet, TikTok, Wise, Mercury, Deel, Replit, Resend, Zoom, Fireflies, Typeform, Canva, Facebook, Google Admin, Hostinger, Google Maps, Retell AI, UptimeRobot, Google Photos, Excel, Cloudflare, Beehiiv, Vapi, 1Password, RunPod, Unipile, Read AI, Figma, Microsoft Teams, Trello, Salesforce, Jira, Confluence, Miro, Coda, Todoist, Basecamp, Wrike, Zendesk, Intercom, Freshdesk, ServiceNow, Gorgias, Mailchimp, ActiveCampaign, MailerLite, SendGrid, Postmark, Customer_io, Gong, Clay, Greenhouse, FreshBooks, Zoho Books, Brex, Ramp, PandaDoc, Google Ads, Microsoft Ads, Pinterest Ads, Semrush, Moz, Mixpanel, Google Tag Manager, Microsoft Clarity, Tableau, Metabase, Snowflake, Google BigQuery, ClickHouse, Wix, Contentful, Sanity, Bubble, Make, GitLab, Bitbucket, Sentry, PagerDuty, Grafana, New Relic, Docker Hub, DigitalOcean, Render, Firebase, Prisma, Algolia, Cloudinary, LaunchDarkly, Postman, Midjourney, Veo, GPT-Image2, NanoBanana, Flux, Kling, MiniMax, Replicate, HuggingFace, DeepSeek, Mistral AI, Ollama, Firecrawl, Eventbrite, Webex, Toggl, Harvest, Tailscale, Mural, Discord, TickTick, Calendly, Attio, Higgsfield, Square, Wispr Flow, Readwise, Bright Data, Google Places, FastMoss, PageSpeed Insights, Kalodata, Cerebras, SerpApi, DataForSEO… Kylon enables them to work better, together.

My conversation with @alexandr_wang, chief AI officer of Meta. 0:00 The vision behind Muse 1:32 Discovering the potential of personal AI agents 3:34 Three years of therapy in one AI session 5:46 Building Muse in 7 months 9:16 Why Meta bet on personal AI 16:51 Mark Zuckerberg's obsession with improving products 20:15 How Meta knew Muse was ready to launch 23:32 Why great products need a single point of view 25:56 Making Meta cool again with memes 34:38 Building trust with AI agents 38:43 The future of human ambition 45:22 The call from Mark Zuckerberg that changed everything 52:13 Leaving Scale AI for Meta 55:19 Rebuilding Meta's AI lab from scratch 58:13 From founder to coach 1:05:55 The future billionaires at Alexandr's 19th birthday party 1:08:04 Diamond mining vs. building skyscrapers 1:09:21 Why Alexandr has more than 200 direct reports Includes paid partnerships.

音视频的实时决策模型， 场景太丰富了，搞了几个场景 demo 让大家感受下。 会带来一波 VoiceAgent 以及 具身智能场景的落地。

🚨 BREAKING: NVIDIA in talks to acquire or acqui-hire Reflection AI, last valued at $25 BILLION >be Reflection AI >backed by the Pentagon and the DOE >nvidia one of its largest shareholders ($800M invested) >Jensen: “we need a domestic open-weight ecosystem” >sign $150M/month compute deal with SpaceX >training on nvidia gb300s at colossus 2 >3 months later: drops "Beam" first open-weight model on par with GLM-5.2 >White House now sees it as the US answer to chinese open-weight labs >deal could close in weeks ITS HAPPENING

Meet Mistral Large 4, aka Le Chonk. • 1T parameters, natively multimodal. 49B active. It is the best open weights model from US or Europe on aggregated benchmarks. • State-of-the-art on critical workloads, including cyber defense, manufacturing and finance and it surpasses closed frontier models on visual grounding. • Forged in Europe end-to-end and is deployable from Europe via our own Mistral Cloud infrastructure. • Available to all via API today. Working with cybersecurity partners privately. Open weights release end of October.