feat：增加了制作视频的功能。

fastapi_app/.env.example

@@ -66,3 +66,11 @@ SEARCH_PROVIDER=serper
 SERPER_API_KEY=your_serper_key_here
 SERPAPI_KEY=your_serpapi_key_here
 BOCHA_API_KEY=your_bocha_key_here
+
+# ============================================
+# Paper2video
+# GUI-Plus（paper2video cursor 等）
+# LivePortrait（数字人）
+# ============================================
+GUI_PLUS_API_KEY=your_dashscope_or_bailian_api_key
+LIVEPORTRAIT_KEY=your_liveportrait_api_key

fastapi_app/config/settings.py

@@ -96,6 +96,10 @@ class AppSettings(BaseSettings):
     LLM_API_KEY: str = ""
     LLM_MODEL: str = "gemini-2.5-flash"
 
+    # Paper2Video: GUI-Plus cursor（阿里云百炼 / DashScope compatible-mode）、LivePortrait（数字人）
+    GUI_PLUS_API_KEY: Optional[str] = None
+    LIVEPORTRAIT_KEY: Optional[str] = None
+
     # Legacy: Local service switches (backward compatibility)
     USE_LOCAL_TTS: int = 0
     TTS_ENGINE: str = "qwen"

fastapi_app/notebook_paths.py

@@ -9,6 +9,7 @@
     │                         /markdown/
     ├── vector_store/
     ├── ppt/{timestamp}/
+    ├── video/{timestamp}/
     ├── mindmap/{timestamp}/
     ├── podcast/{timestamp}/
     └── drawio/{timestamp}/

fastapi_app/routers/kb_outputs_v2.py

@@ -2,7 +2,8 @@
 
 from typing import Any, Dict, List, Optional
 
-from fastapi import APIRouter, Query
+from fastapi import APIRouter, File, Query, UploadFile
+from fastapi.responses import FileResponse
 from pydantic import BaseModel
 
 from fastapi_app.services.output_v2_service import OutputV2Service
@@ -29,6 +30,12 @@ class OutlineRequest(BaseModel):
     api_key: Optional[str] = None
     model: Optional[str] = None
     enable_images: Optional[bool] = None
+    language: Optional[str] = None
+    tts_model: Optional[str] = None
+    tts_voice_name: Optional[str] = None
+    avatar_mode: Optional[str] = None
+    avatar_id: Optional[str] = None
+    avatar_upload_token: Optional[str] = None
 
 
 class SaveOutlineRequest(BaseModel):
@@ -87,6 +94,17 @@ class GenerateOutputRequest(BaseModel):
     model: Optional[str] = None
 
 
+class Paper2VideoWorkflowRequest(GenerateOutputRequest):
+    """paper2video 专用：可选 TTS、语言与数字人（默认读产出项已保存配置）。"""
+
+    tts_model: Optional[str] = None
+    tts_voice_name: Optional[str] = None
+    language: Optional[str] = None
+    avatar_mode: Optional[str] = None
+    avatar_id: Optional[str] = None
+    avatar_upload_token: Optional[str] = None
+
+
 class RegeneratePptPageRequest(GenerateOutputRequest):
     prompt: str
 
@@ -155,10 +173,50 @@ async def create_outline(request: OutlineRequest) -> Dict[str, Any]:
         api_key=request.api_key,
         model=request.model,
         enable_images=request.enable_images,
+        language=request.language,
+        tts_model=request.tts_model,
+        tts_voice_name=request.tts_voice_name,
+        avatar_mode=request.avatar_mode,
+        avatar_id=request.avatar_id,
+        avatar_upload_token=request.avatar_upload_token,
     )
     return {"success": True, "output": item}
 
 
+@router.get("/paper2video/options")
+async def paper2video_options() -> Dict[str, Any]:
+    return {"success": True, **service.get_paper2video_options()}
+
+
+@router.get("/paper2video/preset-asset")
+async def paper2video_preset_asset(
+    kind: str = Query(...),
+    asset_id: str = Query(..., alias="id"),
+) -> FileResponse:
+    path = service.resolve_paper2video_preset_asset(kind=kind, asset_id=asset_id)
+    media = "image/png" if kind == "avatar" else "audio/wav"
+    return FileResponse(path, media_type=media, filename=path.name)
+
+
+@router.post("/paper2video/upload-avatar")
+async def paper2video_upload_avatar(
+    notebook_id: str = Query(...),
+    notebook_title: str = Query(""),
+    user_id: str = Query("local"),
+    email: Optional[str] = Query(None),
+    file: UploadFile = File(...),
+) -> Dict[str, Any]:
+    content = await file.read()
+    payload = service.save_paper2video_avatar_upload(
+        notebook_id=notebook_id,
+        notebook_title=notebook_title,
+        user_id=_effective_user(user_id, email),
+        filename=file.filename or "avatar.png",
+        content=content,
+    )
+    return {"success": True, **payload}
+
+
 @router.put("/{output_id}/outline")
 async def save_outline(output_id: str, request: SaveOutlineRequest) -> Dict[str, Any]:
     item = service.save_outline(
@@ -263,6 +321,50 @@ async def generate_output(output_id: str, request: GenerateOutputRequest) -> Dic
     return {"success": True, "output": item}
 
 
+@router.post("/{output_id}/paper2video/run-subtitle")
+async def paper2video_run_subtitle(output_id: str, request: Paper2VideoWorkflowRequest) -> Dict[str, Any]:
+    item = await service.run_paper2video_subtitle(
+        notebook_id=request.notebook_id,
+        notebook_title=request.notebook_title,
+        user_id=_effective_user(request.user_id, request.email),
+        email=(request.email or request.user_id or "local").strip() or "local",
+        output_id=output_id,
+        api_url=request.api_url,
+        api_key=request.api_key,
+        model=request.model,
+        tts_model=request.tts_model,
+        tts_voice_name=request.tts_voice_name,
+        language=request.language,
+        avatar_mode=request.avatar_mode,
+        avatar_id=request.avatar_id,
+        avatar_upload_token=request.avatar_upload_token,
+    )
+    return {"success": True, "output": item}
+
+
+@router.post("/{output_id}/paper2video/continue-after-edit")
+async def paper2video_continue_after_edit(
+    output_id: str, request: Paper2VideoWorkflowRequest
+) -> Dict[str, Any]:
+    item = await service.run_paper2video_continue_after_edit(
+        notebook_id=request.notebook_id,
+        notebook_title=request.notebook_title,
+        user_id=_effective_user(request.user_id, request.email),
+        email=(request.email or request.user_id or "local").strip() or "local",
+        output_id=output_id,
+        api_url=request.api_url,
+        api_key=request.api_key,
+        model=request.model,
+        tts_model=request.tts_model,
+        tts_voice_name=request.tts_voice_name,
+        language=request.language,
+        avatar_mode=request.avatar_mode,
+        avatar_id=request.avatar_id,
+        avatar_upload_token=request.avatar_upload_token,
+    )
+    return {"success": True, "output": item}
+
+
 @router.post("/{output_id}/pages/{page_index}/regenerate")
 async def regenerate_ppt_page(
     output_id: str,
@@ -318,6 +420,61 @@ async def select_ppt_page_version(
     return {"success": True, "output": item}
 
 
+@router.post("/{output_id}/scenes/{scene_index}/regenerate")
+async def regenerate_video_scene(
+    output_id: str,
+    scene_index: int,
+    request: RegeneratePptPageRequest,
+) -> Dict[str, Any]:
+    item = await service.regenerate_video_scene(
+        notebook_id=request.notebook_id,
+        notebook_title=request.notebook_title,
+        user_id=_effective_user(request.user_id, request.email),
+        email=(request.email or request.user_id or "local").strip() or "local",
+        output_id=output_id,
+        scene_index=scene_index,
+        prompt=request.prompt,
+        api_url=request.api_url,
+        api_key=request.api_key,
+        model=request.model,
+    )
+    return {"success": True, "output": item}
+
+
+@router.post("/{output_id}/scenes/{scene_index}/confirm")
+async def confirm_video_scene(
+    output_id: str,
+    scene_index: int,
+    request: GenerateOutputRequest,
+) -> Dict[str, Any]:
+    item = service.confirm_video_scene(
+        notebook_id=request.notebook_id,
+        notebook_title=request.notebook_title,
+        user_id=_effective_user(request.user_id, request.email),
+        output_id=output_id,
+        scene_index=scene_index,
+    )
+    return {"success": True, "output": item}
+
+
+@router.post("/{output_id}/scenes/{scene_index}/versions/{version_id}/select")
+async def select_video_scene_version(
+    output_id: str,
+    scene_index: int,
+    version_id: str,
+    request: SelectPptPageVersionRequest,
+) -> Dict[str, Any]:
+    item = service.select_video_scene_version(
+        notebook_id=request.notebook_id,
+        notebook_title=request.notebook_title,
+        user_id=_effective_user(request.user_id, request.email),
+        output_id=output_id,
+        scene_index=scene_index,
+        version_id=version_id,
+    )
+    return {"success": True, "output": item}
+
+
 @router.post("/{output_id}/import-source")
 async def import_output_to_source(output_id: str, request: GenerateOutputRequest) -> Dict[str, Any]:
     result = await service.import_output_to_source(