Search by use case, modality, country of origin, or specifications. Discover model family histories, previous version evolution, and direct code integration snippets.
Native models trained on Indic languages with speech robustness and culturally aligned reasoning.
Frontier foundation models leading global benchmarks in coding, RL reasoning, and omni-modal audio/video.
Breakthrough open-weights architectures with Multi-Head Latent Attention and Apache 2.0 1080p video DiT.
Multilingual models, 160ms voice (Moshi), and reigning open image generation models (FLUX.1 Freiburg).
Chronological progression of previous versions, checkpoints, and base architecture milestones.