This is a Next.js project bootstrapped with create-next-app.

Getting Started

Required environment variables:

NEXT_PUBLIC_SUPABASE_URL=
NEXT_PUBLIC_SUPABASE_PUBLISHABLE_KEY=
SUPABASE_SECRET_KEY=
R2_BUCKET_NAME=
R2_ACCESS_KEY_ID=
R2_SECRET_ACCESS_KEY=
R2_ENDPOINT=
TRIGGER_SECRET_KEY=
ASSEMBLYAI_API_KEY=
OPENAI_API_KEY=
OPENAI_WORKER_MODEL=gpt-5.5
OPENAI_REASONING_EFFORT=high
OPENAI_IMAGE_DETAIL=high
OPENAI_VISUAL_ANALYSIS_MAX_OUTPUT_TOKENS=20000
OPENAI_EDIT_PLAN_MAX_OUTPUT_TOKENS=20000
OPENAI_OVERLAY_PLAN_MAX_OUTPUT_TOKENS=10000
OPENAI_INSTRUCTION_DOCUMENT_MAX_OUTPUT_TOKENS=12000
OPENAI_VOICEOVER_SCRIPT_MAX_OUTPUT_TOKENS=12000
OPENAI_TTS_MODEL=gpt-4o-mini-tts
OPENAI_TTS_VOICE=cedar
VIDEO_ANALYSIS_PROVIDER=openai
TWELVELABS_API_KEY=
TWELVELABS_ANALYZE_MODEL=pegasus1.5
GEMINI_API_KEY=
GEMINI_VIDEO_MODEL=gemini-3.5-flash
GEMINI_VIDEO_EVENT_ANALYSIS_ENABLED=false
GEMINI_VIDEO_EVENT_ANALYSIS_REQUIRED=false
VIDEO_STYLE=instruction_overlay

Supabase Auth is used for magic-link sign-in. Add the local and production app URLs to Supabase Auth redirect allow-list, including http://localhost:3000/auth/callback for local development and the production /auth/callback URL for deployment. NEXT_PUBLIC_SITE_URL is optional and is used as a fallback origin for sign-in links when request headers are not available.

For local UI development only, set AUTH_DISABLED_FOR_DEV=true in .env.local to skip magic-link sign-in. This flag is ignored when NODE_ENV=production.

To turn off the app login gate in production, set AUTH_DISABLED=true in the deployment environment. Supabase is still required for database access, but the app will skip Supabase Auth and treat all visitors as a shared bypass user. This makes the app public to anyone with the URL, including access to paid upload and processing services.

ASSEMBLYAI_BASE_URL is optional and defaults to https://api.assemblyai.com. OPENAI_BASE_URL is optional and defaults to https://api.openai.com/v1. OPENAI_WORKER_MODEL is optional and defaults to gpt-5.5. OPENAI_REASONING_EFFORT is optional and defaults to high for worker Responses API calls. Supported values are none, low, medium, high, and xhigh; higher values can improve complex planning quality but usually increase latency and token cost. OPENAI_IMAGE_DETAIL is optional and defaults to high for sampled-frame visual analysis; supported values are low, high, and auto. The OPENAI_*_MAX_OUTPUT_TOKENS settings are optional per-stage limits for OpenAI Responses output, including reasoning tokens, and default to the values shown in the required environment example above. OPENAI_TTS_MODEL is optional and defaults to gpt-4o-mini-tts. OPENAI_TTS_VOICE is optional and defaults to cedar. OPENAI_TTS_INSTRUCTIONS is optional and can provide voice style guidance for the OpenAI speech request. Generated voiceover audio uses OpenAI text to speech. Subtitle timing uses a second AssemblyAI transcription pass over the generated voiceover, so ASSEMBLYAI_API_KEY is required for both source transcription and subtitle timing. OpenAI requires disclosure to end users that they are hearing an AI-generated voice. VIDEO_ANALYSIS_PROVIDER controls the optional whole-video analysis pass. Set it to twelvelabs for paid timestamped action segmentation, gemini for the Gemini whole-video fallback, or openai to use only the existing transcript plus sampled-frame OpenAI visual analysis. TWELVELABS_API_KEY is required when VIDEO_ANALYSIS_PROVIDER=twelvelabs; TWELVELABS_ANALYZE_MODEL is optional and defaults to pegasus1.5. GEMINI_VIDEO_MODEL is optional and defaults to gemini-3.5-flash. GEMINI_VIDEO_EVENT_ANALYSIS_ENABLED controls optional Gemini whole-video event analysis. GEMINI_VIDEO_EVENT_ANALYSIS_REQUIRED defaults to false; when it is false, Gemini failures are logged and the existing transcript plus sampled-frame pipeline continues. VISUAL_FRAME_SAMPLE_INTERVAL_SECONDS and VISUAL_FRAME_SAMPLE_MAX_FRAMES are optional visual-analysis worker settings and default to 3 and 30. VIDEO_STYLE defaults to instruction_overlay, which adds one readable Vidocu-style action caption per selected segment. Set it to voiceover_subtitles to keep the older voiceover/subtitle presentation.

Before running the worker against a fresh database, apply the Supabase migrations in supabase/migrations/ in order. For production Gemini rollout, apply supabase/migrations/20260612000000_add_video_event_analysis_field.sql before deploying app or worker code that reads video_event_analysis_r2_key.

First, run the development server:

npm run dev
# or
yarn dev
# or
pnpm dev
# or
bun dev

Open http://localhost:3000 with your browser to see the result.

You can start editing the page by modifying app/page.tsx. The page auto-updates as you edit the file.

This project uses next/font to automatically optimize and load Geist, a new font family for Vercel.

Learn More

To learn more about Next.js, take a look at the following resources:

• Next.js Documentation - learn about Next.js features and API.
• Learn Next.js - an interactive Next.js tutorial.

You can check out the Next.js GitHub repository - your feedback and contributions are welcome!

Deploy on Vercel

The easiest way to deploy your Next.js app is to use the Vercel Platform from the creators of Next.js.

Check out our Next.js deployment documentation for more details.