100B / 6B (MoE)
MIT
SOTA open-source omni-MLLM built on Ling-2.0 MoE architecture; supports image, text, video, and audio inputs/outputs; expert-level multimodal cognition with vision-to-knowledge synthesis (plant/animal identification, cultural references, artifact analysis); first industry-wide unified audio generation model producing speech, ambient sound effects, and music on same audio track; ultra-low inference frame rate of 3.1Hz for real-time minute-long audio generation; zero-shot voice cloning with natural language control of voice, speed, intonation, volume, emotion, dialect; Continuous Autoregression with Diffusion Transformer (DiT) head; high-dynamic controllable image generation with native multi-task architecture unifying segmentation, generation, and editing; surpasses Gemini 2.5 Pro in some metrics