bytedance

An open-source super agent harness that orchestrates sub-agents, memory, and sandboxes to handle long-horizon tasks that can take minutes to hours. Supports research, coding, and creative tasks with extensible skills and tool integrations.

ByteDance's open-source LLM-based agent for general software engineering tasks — debugging, refactoring, feature work — across multiple model providers.

ByteDance's pioneering native GUI agent — one model interprets screenshots and emits clicks/typing to operate any desktop or web app.

ByteDance's HTML5 video player with a built-in MP4/MSE parser that saves bandwidth via partial loading.

ByteDance's extensible workflow development framework with a built-in canvas — node-based primitives for designing AI workflows.

ByteDance's open-source text-to-speech model — high-quality voice synthesis with multilingual support and voice cloning.

A 3B-active-parameter native unified multimodal model from ByteDance for image and video understanding, generation, and editing.

Sandbox runtime for safely executing LLM-generated code across many languages — useful for code-agent evaluation.

ByteDance's diffusion-based lip-sync model — tame Stable Diffusion to synchronize mouth movements with arbitrary audio.

Batch-processing extension for ComfyUI that lets you sweep parameters across runs to speed up workflow debugging instead of adjusting one knob at a time.

Highly customizable streaming flow builder for React. Compose step-by-step pipelines (LLM chains, ETL, approval flows) in a visual editor.

High-fidelity face-swapping for video using a diffusion transformer, bridging the image-to-video gap from ByteDance research.

ByteDance's identity-preserving photo recrafting — change pose, scene, or style while keeping the same face identity intact.

Universal customization method for both single- and multi-subject image generation, built on diffusion transformers (ICCV 2025).

Unified style- and subject-driven image generation via disentangled and reward learning (CVPR 2026).

Model-based UI testing tool for Android that learns app transitions to surface stability bugs.

NeurIPS 2025 paper code for consistent multi-subject identity and semantic control in diffusion transformers.

SIGGRAPH 2024 portrait animation model with hierarchical motion attention for expressive face reenactment from a driving video.

ByteDance family of large-scale video-language models for high-quality video description and general video understanding.

A simple Go mocking library derived from ByteDance's internal test practices. Patch functions and methods at runtime for unit testing.

Family of large multimodal models from ByteDance for video understanding and editing tasks.

LLM agent that automates environment setup by generating error-free Dockerfiles for any Python repository.

ByteDance's Android native memory leak detector — pinpoints leaks in C/C++ code from production apps.

ComfyUI node implementing HyperLoRA: parameter-efficient adaptive generation for portrait synthesis (CVPR 2025).

Official ComfyUI node for InfiniteYou with FLUX — identity-preserving image generation and personalization (ICCV 2025 Highlight).