Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

LTX-2: Efficient Joint Audio-Visual Foundation Model

About

Recent text-to-video diffusion models can generate compelling video sequences, yet they remain silent -- missing the semantic, emotional, and atmospheric cues that audio provides. We introduce LTX-2, an open-source foundational model capable of generating high-quality, temporally synchronized audiovisual content in a unified manner. LTX-2 consists of an asymmetric dual-stream transformer with a 14B-parameter video stream and a 5B-parameter audio stream, coupled through bidirectional audio-video cross-attention layers with temporal positional embeddings and cross-modality AdaLN for shared timestep conditioning. This architecture enables efficient training and inference of a unified audiovisual model while allocating more capacity for video generation than audio generation. We employ a multilingual text encoder for broader prompt understanding and introduce a modality-aware classifier-free guidance (modality-CFG) mechanism for improved audiovisual alignment and controllability. Beyond generating speech, LTX-2 produces rich, coherent audio tracks that follow the characters, environment, style, and emotion of each scene -- complete with natural background and foley elements. In our evaluations, the model achieves state-of-the-art audiovisual quality and prompt adherence among open-source systems, while delivering results comparable to proprietary models at a fraction of their computational cost and inference time. All model weights and code are publicly released.

Yoav HaCohen, Benny Brazowski, Nisan Chiprut, Yaki Bitterman, Andrew Kvochko, Avishai Berkowitz, Daniel Shalem, Daphna Lifschitz, Dudu Moshe, Eitan Porat, Eitan Richardson, Guy Shiran, Itay Chachy, Jonathan Chetboun, Michael Finkelson, Michael Kupchick, Nir Zabari, Nitzan Guetta, Noa Kotler, Ofir Bibi, Ori Gordon, Poriya Panet, Roi Benita, Shahar Armon, Victor Kulikov, Yaron Inger, Yonatan Shiftan, Zeev Melumian, Zeev Farbman• 2026

Related benchmarks

TaskDatasetResultRank
Video Generationshort videos 81-frames 240 prompts
Total Score6
38
Video ReasoningVBVR-Bench
Overall Accuracy31.3
11
Video ReasoningVBVR-Bench In-Domain
Average Score32.9
11
Video ReasoningVBVR-Bench Out-of-Domain
Average Score0.297
11
Joint audio-video generationIdentity-aware T2AV (test)
AES0.538
7
Audio-visual generationVerse-Bench (All subsets)
IS (Score)3.066
7
Audio-visual generationVerse-Bench multi-speaker
cpCER22
6
Audio-visual generationVerse-Bench (set3)
DNSMOS3.635
6
Motion-conditioned Audio-Video GenerationAudio-Video Generation Evaluation Set
AS4.31
5
Joint audio-video generationHARD (test)
Sync-C5.47
4
Showing 10 of 14 rows

Other info

GitHub

Follow for update