Fun-ASR-Nano deployment hub: vLLM, GGUF, OpenAI API, and containers

[LauraGPT]

Deployment routes for Fun-ASR-Nano

The new FunASR deployment hub now gives Fun-ASR-Nano users one maintained place to choose between GPU throughput, a portable standalone runtime, and an application-facing API.

• NVIDIA GPU and batch throughput: vLLM guide
• Linux, macOS, Windows, CPU, CUDA, or Vulkan: llama.cpp / GGUF guide
• Existing OpenAI SDKs and agents: OpenAI-compatible private API
• Containers and Kubernetes: deployment guide
• Capacity, security, and rollback review: production checklist

Each page identifies the supported model/runtime combination, exact launch and smoke commands, operational boundaries, known limitations, and source evidence. Benchmark records explicitly qualify missing hardware or timing details rather than presenting them as production capacity promises.

The full release verification and deployment matrix are recorded in the canonical FunASR announcement.

中文和英文页面都已上线：中文 · English