Skip to content
IFAI

微软

Company6 stories

  1. Open sourceReach 48

    Microsoft open-sources Mage Flow for instant image generation and editing

    Microsoft just released Mage Flow, an open-source 4B-parameter model that handles both text-to-image generation and instruction-based editing. It features native ControlNet capabilities and multilingual text rendering, while a 4-step Turbo variant spits out 1024px images in under 0.6 seconds on a single A100. At 17.5GB, it brings fast, interactive image manipulation directly to mid-to-high-end GPUs.

  2. Open sourceReach 48

    Microsoft turns tutorial videos into executable AI agent skills

    Microsoft just open-sourced Resource2Skill, a framework that distills human tutorial videos, articles, and code into executable agent skills. Instead of manual prompt writing, it builds a multimodal skill library hooked up to MCP servers, enabling agents to generate real Web pages, PowerPoint decks, and Blender scenes.

  3. Open sourceReach 48

    Microsoft open-sources Lens, a 3.8B image model with sub-second generation

    Microsoft just released weights for Lens, a 3.8B text-to-image model that rivals larger 6B systems at a fraction of the training cost. Its distilled Turbo variant generates images in just 4 steps and 0.84 seconds on an H100, supporting resolutions up to 1440x1440. It brings high-end, lightning-fast generation directly to independent developers.

  4. Open sourceReach 48

    Microsoft open-sources VibeVoice-ASR for 60-minute single-pass audio

    Microsoft just open-sourced VibeVoice-ASR, a 9B speech model capable of transcribing up to 60 minutes of continuous audio in a single pass. Instead of slicing audio into chunks, it unifies transcription, speaker diarization, and timestamps into one end-to-end generation task. It basically solves context fragmentation for long meetings and podcasts.

  5. AI toolsReach 68

    Microsoft launches Copilot Vision preview in Edge to browse the web with you

    Microsoft just launched a preview of Copilot Vision inside Edge, allowing the AI to see what is on your screen and talk with you in real time. Browsing sessions are opt-in and deleted immediately afterward to address privacy concerns. It is currently rolling out to select Copilot Pro subscribers in the US via Copilot Labs.

  6. AI toolsReach 70

    Microsoft gives Copilot eyes, a voice, and deeper thinking

    Microsoft just gave Copilot its biggest overhaul yet, introducing a natural Voice mode, a Vision feature that sees what you're browsing, and Think Deeper for complex multi-step reasoning. It also launched a morning audio briefing powered by licensed publishers like Reuters and FT. Vision and deeper reasoning are rolling out first via Copilot Labs to select users.