LightX2V AI Video Generation

Use the lightx2v CLI to submit LightX2V tasks, poll results, create/run workflows, synthesize speech, and manage cloned voices.

Start Here

1. Authenticate: use LIGHTX2V_API_KEY, LIGHTX2V_CLOUD_API_KEY, or lightx2v login.
2. Set backend if needed: LIGHTX2V_BASE_URL=https://x2v.light-ai.top or a local server such as http://127.0.0.1:8081.
3. Discover live models before guessing names: lightx2v models --json.
4. Prefer direct lightx2v run for one model task. Use lightx2v workflow ... for multi-node graphs, saved user workflows, or generated TTS -> digital-human pipelines.
5. Before creating or editing workflow JSON, read references/workflow-nodes.md for node ids, ports, data fields, and graph patterns.
6. Before discovering inputs, running, monitoring, or cancelling a saved workflow, read references/workflow-api.md.
7. Before using TTS or voice clone, read references/voice-tts.md for voice discovery, parameters, clone commands, and workflow tts nodes.
8. For multi-person or multi-role S2V, read references/s2v-multi.md before preparing masks, role tracks, or a request. Its user-confirmation gate is mandatory.
9. Never print API keys, bearer tokens, or long signed media URLs in shared output.

Install: