ByteDance’s video editor CapCut just announced a ChatGPT plugin that turns raw footage into a rough first cut through simple prompts. Creators can keep refining the timeline in chat—tweaking shot lengths and subtitles—or adjust things manually. It’s set to roll out for testing in the coming months.
ByteDance introduced DMAD, an adversarial distillation method for rapid visual generation. By training a student model to match real data distributions using a shared-backbone discriminator, it compresses 20-to-30-step video diffusion models like MiniMax-H3 down to four denoising steps while preserving fidelity.
Researchers just open-sourced PDMD, a video distillation technique that shrinks MiniMax-H3 generation from 30 steps down to just four. By curbing misleading critic feedback during training, it keeps motion smoother and visual details sharp instead of muddying the frame. Fast open-source video generation is getting shockingly usable.
ByteDance researchers just introduced Lucida, a pipeline that turns indoor video captures into interactive 3D simulations. Instead of outputting a single frozen mesh, it builds clean 3D assets for each detected object and uses a vision-language agent to place them precisely back into position. Why it matters: It gives robotics and embodied AI fully manipulable digital twins of real-world spaces.
ByteDance just open-sourced Bernini 2, an omnimodal video pipeline that lets you edit existing footage using natural language, from swapping backgrounds to inserting new characters. It pairs an MLLM planner with a diffusion renderer for fine-grained control, but weighing in at 180GB for the model alone, running it locally will be a heavy lift.
ByteDance just launched Seedance 2.5, a video model that produces 30-second audio-video clips in a single pass with multi-round extension for longer narratives. It can ingest up to 30 images, 10 videos, and 10 audio clips simultaneously as reference inputs, alongside timestamp-level editing and clay render guidance. It is live now on Jimeng AI and Doubao Pro.
ByteDance just launched Seedream 5.0 Pro, an image model tailored for real design workflows instead of random generation. It tackles high-density infographics with complex typography, separates elements into layers, and executes edits based on spatial sketches and handwritten annotations. It basically turns image generation into a controllable creative collaborator.
ByteDance just launched the Doubao Seed 2.1 model family, introducing Pro and Turbo versions built specifically as productivity agents. Alongside strong coding and multimodal chart analysis, it can accurately overhaul PPT templates and generate websites from rough sketches. It marks ByteDance's clearest push yet to shift AI from a simple chatbot into a reliable digital coworker.
ByteDance just launched Seed-Audio 1.0, an all-in-one audio model capable of generating multi-character dialogue, emotional voice acting, sound effects, and background music in a single pass. It uses audio references to keep voices consistent across long podcasts or shows, and is currently rolling out via API.
ByteDance just open-sourced Bernini, a unified video framework pairing an MLLM semantic planner with a DiT renderer. It handles generation and editing across text, images, and video in a single pipeline, putting AI video closer to real directing. Code and model weights are now live on GitHub and Hugging Face.
ByteDance just open-sourced Lance, a lightweight unified multimodal model with 3B active parameters that handles understanding, generation, and editing for both images and videos. Basically, it allows users to continuously tweak video clips through multi-turn dialogue, though running it locally requires a beefy 40GB GPU.
ByteDance just launched Seed 2.0, an updated suite of general-purpose agent models spanning Pro, Lite, and Mini sizes. Pro tackles long-chain reasoning in messy real-world workflows, Mini handles high-concurrency batch tasks, and Lite natively fuses audio, video, and text understanding. Why it matters: Frontier labs are shifting from isolated benchmarks to long-horizon agent execution in enterprise production.
ByteDance and CUHK researchers just introduced OmniShow, an end-to-end video model built specifically for human-object interactions. It accepts product images, character photos, audio tracks, and pose skeletons all at once, generating up to 10-second marketing clips with synced lip-sync and hand movements. The code is currently under internal review on GitHub ahead of an open-source release.
ByteDance just introduced DreamLite, a compact 0.39B diffusion model that handles both text-to-image generation and image editing directly on smartphones. It can generate a 1024×1024 image in around 3 seconds on an iPhone, swap styles, and remove objects completely offline. While fine details still lag behind massive server models, it brings fast, private image manipulation right to local devices.
ByteDance just open-sourced Lark CLI under the MIT license, specifically designed for human developers and AI agents. It spans 18 enterprise domains—including Docs, Base, and Calendar—with over 200 curated commands. Why it matters: popular AI coding tools like Claude Code can now hook into and operate your entire Lark workspace with zero setup.
ByteDance just open-sourced DeerFlow 2.0, an MIT-licensed agent orchestration harness designed for complex, multi-hour jobs. Instead of acting as a thin wrapper, a lead agent breaks tasks down and spins up isolated sub-agents inside Docker containers with dedicated filesystems and persistent memory. It works with local or cloud models like DeepSeek and GPT-4.
ByteDance just open-sourced Helios, a 14B video generation model that hits 19.5 FPS on a single Nvidia H100 GPU. It can churn out minute-long videos supporting text, image, and video inputs without quality drifting. Why it matters: bringing near-real-time speeds to minute-scale generation gets open-source video much closer to interactive production.
ByteDance and Stanford researchers just open-sourced Spectrum, a training-free framework that accelerates image and video diffusion models by up to 4.7x. By forecasting future step trajectories with Chebyshev polynomials, it lets models like FLUX.1 and Hunyuan Video generate outputs much faster with virtually no drop in quality.
ByteDance and Tsinghua just introduced DreamID-Omni, an AI model that turns reference portraits, text, and voice clips into talking videos. It eliminates lip-sync confusion in multi-person scenes and supports full video face-and-voice editing, with open-source code expected in March.
ByteDance researchers introduced a spatial reward modeling framework alongside SpatialScore, fixing a stubborn flaw where image generators mix up object positions. By training on over 80,000 spatial preference pairs, the method helps models like Flux follow complex layout prompts. The team plans to open-source the dataset and weights soon.
