Audio GGUF models
1518 GGUF models: speech, music and audio models, which are not language models. Sizes run from 1M to 35.9B parameters. The longest trained context is 1024K tokens.
About Audio GGUF models
1,518 GGUF models are filed under Audio: speech, music and audio models, which are not language models. The most downloaded are YuE2 audio tokenizer, the v9 head of Mothersuperior, on MERT-v2-FullSong, Parakeet Unified EN 0.6B and nemotron 3.5 asr streaming 0.6b. Parameter counts run from 1M to 35.9B, and the longest trained context window is 1024K tokens.
These are speech, music or audio models, not language models, and they need a runtime built for them. The memory estimates on this site are rough for them. The families you will see most are Qwen (168), Parakeet (153) and Whisper (121).
Frequently asked questions
What are Audio GGUF models?
Speech, music and audio models, which are not language models. There are 1,518 of them on this site.
Which Audio GGUF model should I start with?
The most downloaded is YuE2 audio tokenizer, the v9 head of Mothersuperior, on MERT-v2-FullSong. Choose the size that fits your GPU, then a Q4_K_M quant unless you have memory to spare.