OpenTalking v0.1.0

OpenTalking v0.1.0

OpenTalking v0.1.0 is the first GitHub Release for the real-time digital-human orchestration layer. It packages the FastAPI API, async worker, React web console, model-provider registry, and documentation-backed deployment paths.

Artifacts

• Python source distribution and wheel are attached to this release.
• GHCR images:
  • ghcr.io/datascale-ai/opentalking-api:v0.1.0
  • ghcr.io/datascale-ai/opentalking-worker:v0.1.0
  • ghcr.io/datascale-ai/opentalking-web:v0.1.0

Install

python -m pip install opentalking-0.1.0-py3-none-any.whl
opentalking-doctor --help

For source installs and model-specific setup, follow the documentation site. Model weights are not bundled in the Python artifacts or Docker images.

Docker

docker pull ghcr.io/datascale-ai/opentalking-api:v0.1.0
docker pull ghcr.io/datascale-ai/opentalking-worker:v0.1.0
docker pull ghcr.io/datascale-ai/opentalking-web:v0.1.0

The release images package the OpenTalking orchestration services. Real talking-head inference still requires a configured local model backend or an OmniRT-compatible remote runtime.

Compatibility Notes

• Python 3.10+ is supported; CI validates Python 3.11.
• Frontend builds use Node.js 20.
• Docker images do not include large model weights.
• GPU/NPU runtime compatibility depends on the selected model backend and host driver stack.

What's Changed

• Split runtime core from MuseTalk/Wav2Lip changes by @XX123122 in #4
• feat: 百炼多线路TTS/STT、字幕同步、TTS开场白、idle视频、FlashTalk队列调度、新增录制和上传音频功能 by @pb19834141522-a11y in #8
• fix: 修复CI告警 by @zyairehhh in #9
• 文档与quickstart逻辑优化 by @zyairehhh in #10
• 增加 README 演示视频展示区 by @pb19834141522-a11y in #11
• feat: add musetalk and wav2lip runtime support by @XX123122 in #13
• Fix TTS error surfacing and FlashTalk A/V sync by @cwang0810 in #14
• Improve quickstart defaults and avatar asset alignment by @cwang0810 in #17
• Refactor model defaults into YAML configs by @cwang0810 in #18
• [feature] Improve web workbench UI feedback by @cwang0810 in #19
• [feature] Refine FlashHead adapter integration by @cwang0810 in #20
• avatar selection layout by @zyairehhh in #21
• Add QuickTalk model adapter by @cwang0810 in #22
• Refactor/architecture v2 by @kero-ly in #27
• Fix：修复昇腾环境下安装依赖的问题;将env中视频处理参数放置对应启动脚本 by @charm-ch in #29
• bugfix & update readme by @kero-ly in #30
• Decouple synthesis model backends by @cwang0810 in #31
• Docs: restructure model backend documentation by @cwang0810 in #32
• Docs: rebuild model deployment docs and add Pages publishing by @cwang0810 in #33
• docs: fix install and deployment guidance by @cwang0810 in #34
• docs: streamline README entry points by @cwang0810 in #35
• docs: restore community section position by @cwang0810 in #37
• ci: skip docs pages deploy outside upstream repo by @zyairehhh in #39
• 适配MuseTalk by @charm-ch in #40
• feat: route QuickTalk through OmniRT audio2video by @zyairehhh in #41
• feat:Add MuseTalk media backpressure for MuseTalk WebRTC playback by @charm-ch in #42
• docs: restructure documentation navigation by @cwang0810 in #44
• feat: add realtime FasterLivePortrait support by @zyairehhh in #45
• readme update & support local quicktalk, wav2lip by @kero-ly in #43
• docs: add agent collaboration guide by @zyairehhh in #46
• feat: align local wav2lip runtime parity by @zyairehhh in #48
• feat: align wav2lip and quicktalk asset handling by @zyairehhh in #49
• Improve Wav2Lip avatar playback and per-asset preload cache by @lyfics in #47
• Add OpenTalking E2E benchmark tooling by @charm-ch in #50
• Add Cantonese Qwen voices and avatar updates by @lyfics in #51
• feat: align avatar cache prewarm flow by @zyairehhh in #52
• website structure update by @kero-ly in #54
• Feat/docs update by @lyfics in #55
• refactor: unify local and omnirt audio2video flow by @zyairehhh in #57
• docs: update QuickTalk weight download instructions by @zyairehhh in #58
• feat: add local musetalk backend support by @charm-ch in #59
• Add local STT/TTS pipeline with provider-scoped runtime config by @zyairehhh in #61
• docs: add benchmark guide, WSL2 fix, test results, and windows deploy… by @lyfics in #62
• Docs update from latest main by @lyfics in #64
• docs: refresh README roadmap by @zyairehhh in #65
• feat: add FasterLivePortrait video clone workflow by @zyairehhh in #70
• feat: add web recording exports and asset library by @zyairehhh in #71
• homepage design by @kero-ly in #72
• docs: reorganize model deployment guides by @zyairehhh in #73
• docs: add WebUI video workflows by @zyairehhh in #75
• feat: support FasterLivePortrait video creation by @zyairehhh in #76
• docs: add Atlas Cloud as an OpenAI-compatible LLM provider option by @lucaszhu-hue in #74
• feat: add agent knowledge and audio video exports by @lyfics in #78
• docs: reorganize model deployment guides by @zyairehhh in #79
• feat: add openai-compatible audio providers by @zyairehhh in #80
• docs: V100部署指南与避坑 by @zhou693y in #77
• feat: adapt knowledge base asset workflow by @lyfics in #81
• add en page, optimize homepage for deploy by @kero-ly in #82
• Improve MuseTalk deployment setup by @zyairehhh in #83
• homepage video sound & bugfix by @kero-ly in #84
• docs: update README demo videos by @lyfics in #90
• switch to tencent source by @kero-ly in #89
• docs: update README A3 demo video by @lyfics in #91
• feat: add Persona Package support by @cwang0810 in #87
• docs: update C2 demo video link by @lyfics in #93
• add homepage url router by @kero-ly in #94
• feat: add LightRAG knowledge retrieval by @lyfics in #92
• feat: add character memory workflow by @charm-ch in #69
• feat: improve IndexTTS and QuickTalk video creation by @zyairehhh in #95
• fix quicktalk local assets and support QuickTalk on Apple Silicon by @zyairehhh in #98
• feat: add flashtalk reference video generation by @charm-ch in #97
• docs:add WSL2 network mode selection guide for Windows deployment by @Le0der in #99
• add homepage analyze by @kero-ly in #96

New Contributors

• @XX123122 made their first contribution in #4
• @pb19834141522-a11y made their first contribution in #8
• @cwang0810 made their first contribution in #14
• @kero-ly made their first contribution in #27
• @charm-ch made their first contribution in #29
• @lyfics made their first contribution in #47
• @lucaszhu-hue made their first contribution in #74
• @zhou693y made their first contribution in #77
• @Le0der made their first contribution in #99

Full Changelog: https://github.com/datascale-ai/opentalking/commits/v0.1.0