zai-org/GLM-4-Voice

Health v2 · maintenance only

Flags: no_releases

Adoption not part of the score

3222 stars · 286 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-29, confidence not recorded

GLM-4-Voice is an end-to-end bilingual (Chinese/English) speech dialogue model from Zhipu AI, built on GLM-4-9B with a speech tokenizer and a flow-matching speech decoder. It understands and generates voice directly, supports real-time low-latency voice chat, and can adjust emotion, tone, speed, and dialect per user instructions.

Use cases

• build a real-time voice chatbot that speaks Chinese and English
• generate speech responses with controllable emotion and speaking speed
• run an end-to-end speech-to-speech conversation model locally
• transcribe and understand spoken audio into discrete tokens
• deploy a web demo for voice conversations with text and audio replies

When to choose

• you need low-latency end-to-end voice dialogue rather than separate ASR + LLM + TTS pipelines
• you want a model that follows voice instructions to change emotion, tone, speed, or dialect
• you need strong Chinese and English speech understanding and generation in one model

When to avoid

• you need production-grade multilingual support beyond Chinese and English
• you lack GPU resources to run a 9B-parameter model and its decoder
• you only need simple text-to-speech or speech-to-text without conversational intelligence

Facets

library · maturity active

speech-recognition tts llm-inference chatbot audio-processing speech-processing large-language-models artificial-intelligence python cross-platform voice-assistant end-to-end-speech speech-tokenizer streaming-inference chinese-english web-demo glm natural-language-processing gpu

Member repositories

For agents

markdown · JSON · MCP: product_card(name="zai-org/GLM-4-Voice")