AI Briefingaudio.cpp 0.7 Supports 62 Model Families
Key point
The local audio AI framework audio.cpp 0.7 has been released, supporting 62 model families and an Arena UI for comparison.
Details
Version 0.7 of the local audio AI framework audio.cpp has been released, now supporting a total of 62 model families and over 85 model variants.
The key feature of this update is the Arena UI, which allows users to apply the same input to multiple local models or GGUF variants and compare the generated outputs side-by-side. This is useful for selecting the optimal model without writing scripts.
Key New Models
- TTS/Voice: FireRedTTS3, MagpieTTS, PersonaPlex, F5-TTS/Habibi, MOSS VoiceGenerator, DotTTS Edit
- ASR/Understanding: FireRedAudio, IBM Granite Speech 5.0 TurboCTC, MMS Forced Aligner
- Voice Conversion: MeanVC2
- Music/Generation: MiniMax Music 3, MiDashengLM-Gen, ControlFoley (experimental), ACE-Step 1.5 XL
- Tools: AudioSR
Hardware and Deployment
Contributor testing results confirmed that 40 out of 40 model families work correctly on NVIDIA Jetson Orin NX 16GB, and 34 work on Orin Nano 8GB.
Deployable prebuilt binaries support Windows (CPU, Vulkan, CUDA 12.4/13.3), Ubuntu x64 (CPU, Vulkan), and macOS (arm64 Metal, x64 CPU) environments.
This summary was generated automatically by AI. Check the original for the author's claims and context. Copyright belongs to the original author.
Our guide explains how the AI works. Report summary errors, attribution issues, or removal requests via Contact.
Related
Jevper Released: A Jev Classification Wrapper for OpenAI-Compatible Models with Automatic Logprobs Detection