Model Hub

Chatterbox — Resemble AI open-source text-to-speech model with voice cloning capability and natural prosody. MIT license.

TinyLlama 1.1B chat — greedy decode contract. Published under its Hugging Face repository name (TinyLlama/TinyLlama-1.1B-Chat-v1.0); the earlier record TinyLlama/TinyLlama-1.1B-Chat carried an invented name and is hidden in favour of this one.

Whisper Large-V2 — OpenAI's multilingual speech-to-text model for robust transcription and translation across 99 languages, resilient to accents and background noise. Relabeled 2026-08-26: this artifact was previously listed as Large-V3; fingerprint verification (80 mel bins, 99 language tokens) identified it as large-v2-class. MIT license.

Whisper Large-V3 Turbo — OpenAI's speed-optimized speech-to-text model, a fine-tuned variant of Whisper large-v3 with a pruned decoder (4 layers) that delivers much faster multilingual transcription at near-identical accuracy. MIT license.

Text-to-video diffusion, 1.3B parameters, 480x832 clips.

SANA 1600M Multilingual — NVIDIA Labs efficient text-to-image diffusion model with multilingual prompt support (including Chinese and emoji), producing high-resolution images via a linear-attention transformer and deep-compression autoencoder.

PixArt-Σ XL — diffusion-transformer text-to-image model that uses a weak-to-strong training scheme to generate high-fidelity 1024px images (scalable toward 4K) with strong prompt alignment and efficient compute.

Text-to-video diffusion transformer, 2B parameters, 720x480 clips with classifier-free guidance.

Dual-expert image-to-video diffusion, two 14B denoisers with boundary switching, 480P and 720P clips.

DeepSeek-Coder-V2-Lite-Instruct — 15.7B-A2.4B code specialist; MLA attention (kv_lora 512, qk 128+64, v 128) + YaRN-mscale, MoE 64+2 experts top-6

PixArt-α XL — efficient diffusion-transformer text-to-image model producing photorealistic 1024px images with strong prompt following, trained at a small fraction of the cost of comparable diffusion models.

Qwen3-VL 30B-A3B Thinking — text, image and video understanding; 128-expert MoE, DeepStack vision, interleaved M-RoPE; multi-GPU sharded