No matches. Try a model format, tool name or error.
On this page
What are the best open speech & audio models?
Compare the speech and audio models below. Read the model's description to check what it is designed to do; this list does not measure quality or show what works in your app.
License terms and download availability are separate. Read the exact upstream license before using a model.
Answer response assembled as of . This dates the returned rows, not an upstream model update or a successful model run. Check each linked file record for current availability.
142 hosted files. Inspect the file record for current availability. 280 files listed upstream.
File evidence
No exact file name or SHA-256 is recorded in this answer row. Open the file record for current file-level evidence; unknown availability is unconfirmed.
Why listed
The catalog lists it as an audio model. Recorded task: image text to video. Listed formats: safetensors. This listing does not measure quality or prove runtime support.
Model from Hugging Face for image text to text: HauhauCS/Gemma-4-E4B-Uncensored-HauhauCS-Aggressive
License
gemma. Commercial use is permitted with obligations: Gemma terms: commercial use allowed subject to Google's prohibited-use policy. License details
Hosting
12 hosted files. Inspect the file record for current availability. 14 files listed upstream.
File evidence
No exact file name or SHA-256 is recorded in this answer row. Open the file record for current file-level evidence; unknown availability is unconfirmed.
Why listed
The catalog lists it as an audio model. Recorded task: image text to text. Listed formats: gguf. Recorded license: gemma. This listing does not measure quality or prove runtime support.
Model from Hugging Face for audio to audio: nvidia/bigvgan_v2_22khz_80band_256x · PyTorch
License
mit. Permissive open license: commercial use, modification, and redistribution are permitted. License details
Hosting
6 hosted files. Inspect the file record for current availability. 31 files listed upstream.
File evidence
No exact file name or SHA-256 is recorded in this answer row. Open the file record for current file-level evidence; unknown availability is unconfirmed.
Why listed
The catalog lists it as an audio model. Recorded task: audio to audio. Recorded license: mit. This listing does not measure quality or prove runtime support.
Model from Hugging Face for any to any: Qwen/Qwen3-Omni-30B-A3B-Instruct · transformers
License
apache-2.0. Permissive open license: commercial use, modification, and redistribution are permitted. License details
Hosting
19 hosted files. Inspect the file record for current availability. 25 files listed upstream.
File evidence
No exact file name or SHA-256 is recorded in this answer row. Open the file record for current file-level evidence; unknown availability is unconfirmed.
Why listed
The catalog lists it as an audio model. Recorded task: any to any. Listed formats: safetensors. Recorded license: apache-2.0. This listing does not measure quality or prove runtime support.
Model from Hugging Face for automatic speech recognition: handy-computer/Voxtral-Mini-4B-Realtime-2602-gguf · transcribe.cpp
License
apache-2.0. Permissive open license: commercial use, modification, and redistribution are permitted. License details
Hosting
8 hosted files. Inspect the file record for current availability. 8 files listed upstream.
File evidence
No exact file name or SHA-256 is recorded in this answer row. Open the file record for current file-level evidence; unknown availability is unconfirmed.
Why listed
The catalog lists it as an audio model. Recorded task: automatic speech recognition. Listed formats: gguf. Recorded license: apache-2.0. This listing does not measure quality or prove runtime support.
Model from Hugging Face for automatic speech recognition: facebook/seamless-m4t-v2-large · transformers
License
cc-by-nc-4.0. cc-by-nc-4.0: commercial use is not permitted; non-commercial use and adaptations require attribution. License details
Hosting
10 hosted files. Inspect the file record for current availability. 18 files listed upstream.
File evidence
No exact file name or SHA-256 is recorded in this answer row. Open the file record for current file-level evidence; unknown availability is unconfirmed.
Why listed
The catalog lists it as an audio model. Recorded task: automatic speech recognition. Listed formats: safetensors. Recorded license: cc-by-nc-4.0. This listing does not measure quality or prove runtime support.
Model from Hugging Face for text to audio: facebook/musicgen-small · transformers
License
cc-by-nc-4.0. cc-by-nc-4.0: commercial use is not permitted; non-commercial use and adaptations require attribution. License details
Hosting
7 hosted files. Inspect the file record for current availability. 13 files listed upstream.
File evidence
No exact file name or SHA-256 is recorded in this answer row. Open the file record for current file-level evidence; unknown availability is unconfirmed.
Why listed
The catalog lists it as an audio model. Recorded task: text to audio. Listed formats: safetensors. Recorded license: cc-by-nc-4.0. This listing does not measure quality or prove runtime support.
1 hosted file. Inspect the file record for current availability. 12 files listed upstream.
File evidence
No exact file name or SHA-256 is recorded in this answer row. Open the file record for current file-level evidence; unknown availability is unconfirmed.
Why listed
The catalog lists it as an audio model. Recorded task: text to video. Listed formats: gguf. This listing does not measure quality or prove runtime support.
Model from Hugging Face for automatic speech recognition: Oriserve/Whisper-Hindi2Hinglish-Prime · transformers
License
apache-2.0. Permissive open license: commercial use, modification, and redistribution are permitted. License details
Hosting
14 hosted files. Inspect the file record for current availability. 26 files listed upstream.
File evidence
No exact file name or SHA-256 is recorded in this answer row. Open the file record for current file-level evidence; unknown availability is unconfirmed.
Why listed
The catalog lists it as an audio model. Recorded task: automatic speech recognition. Listed formats: safetensors. Recorded license: apache-2.0. This listing does not measure quality or prove runtime support.
1 hosted file. Inspect the file record for current availability. 5 files listed upstream.
File evidence
No exact file name or SHA-256 is recorded in this answer row. Open the file record for current file-level evidence; unknown availability is unconfirmed.
Why listed
The catalog lists it as an audio model. Recorded task: automatic speech recognition. Listed formats: gguf. This listing does not measure quality or prove runtime support.
1 hosted file. Inspect the file record for current availability. 8 files listed upstream.
File evidence
No exact file name or SHA-256 is recorded in this answer row. Open the file record for current file-level evidence; unknown availability is unconfirmed.
Why listed
The catalog lists it as an audio model. Recorded task: automatic speech recognition. Listed formats: gguf. This listing does not measure quality or prove runtime support.
Model from Hugging Face for text to audio: facebook/musicgen-melody · transformers
License
cc-by-nc-4.0. cc-by-nc-4.0: commercial use is not permitted; non-commercial use and adaptations require attribution. License details
Hosting
9 hosted files. Inspect the file record for current availability. 14 files listed upstream.
File evidence
No exact file name or SHA-256 is recorded in this answer row. Open the file record for current file-level evidence; unknown availability is unconfirmed.
Why listed
The catalog lists it as an audio model. Recorded task: text to audio. Listed formats: safetensors. Recorded license: cc-by-nc-4.0. This listing does not measure quality or prove runtime support.
It depends on your task, available memory, file format and license. The records below match this guide’s filters. Check the linked file evidence before choosing: a catalog match does not prove runtime support, safety or answer quality.
An answer pack is a question-specific filter over catalog records. It is not a benchmark, safety review or guarantee that a model will run.