pixeltable-new

Create a new Pixeltable project in one command.

How to use

Install uv following their guide for your system.

Run:

uvx pixeltable-new myapp

This will create a new project myapp with a Pixeltable schema and declarative serving config.

Enter the directory:

cd myapp

Install dependencies and run:

uv sync
uv run python schema.py
uv run pxt serve pipeline

Open your browser and go to http://localhost:8000/docs to see your API docs.

Application Templates

Vertical apps that each build on a structural pattern — so you already know how to run and deploy them:

uvx pixeltable-new --template knowledge-base my-kb             # web UI + API
uvx pixeltable-new --template chat-agent my-agent              # web UI + API
uvx pixeltable-new --template audio-transcription my-podcast   # web UI + API
uvx pixeltable-new --template full-stack-showcase my-sitewatch # web UI + API (complete reference app)
uvx pixeltable-new --template video-search my-video-app        # API only
uvx pixeltable-new --template media-indexing my-pipe           # API + batch
uvx pixeltable-new --template image-dataset my-dataset         # API + batch

Template	Pattern	What you get
knowledge-base	serving + backend	Unified search + RAG Q&A across docs, images, video, audio. schema.py + app.py + web UI
chat-agent	serving + backend	Persistent agent with durable memory, tools, MCP. schema.py + app.py + web UI
audio-transcription	serving + backend	Audio/podcast transcription, summarization, semantic search. schema.py + app.py + web UI
full-stack-showcase	serving + backend	Complete reference app: Gemini + DETR + Whisper, React UI, dashboard. schema.py + app.py + routers/ + frontend/
video-search	serving	Declarative video pipeline: frames, transcription, detection, temporal search. Pure schema.py. Run: pxt serve videointel
media-indexing	batch	Ingest from S3, process all modalities, export to your DB. schema.py + pipeline.py
image-dataset	batch	Auto-annotate, curate, version, export to PyTorch. schema.py + export.py

Structural Patterns

API/pipeline scaffolds for when you want to wire Pixeltable into your own architecture:

uvx pixeltable-new myapp --serving    # Declarative API from TOML config (default)
uvx pixeltable-new myapp --backend    # FastAPI API scaffold (headless)
uvx pixeltable-new myapp --batch      # Batch processing script

Pattern	What you get	Run with
--serving (default)	schema.py + pyproject.toml routes	pxt serve <service-name>
--backend	FastAPI API scaffold + Pixeltable schema + routers	uvicorn main:app --reload
--batch	Ingest script + export_sql	python pipeline.py

Discovery

uvx pixeltable-new --list    # show all patterns + templates

Legacy template names (deprecated since 0.4.2, still work): video-intel → video-search, multimodal-rag → knowledge-base, agent → chat-agent, audio-intel → audio-transcription, content-pipeline → media-indexing, data-lab → image-dataset. Prefer the canonical names above.

video-search quickstart after scaffolding:

cd my-video-app && uv sync && uv run python schema.py && uv run pxt serve videointel

All content is fetched from the Pixeltable Starter Kit. For the full reference with Docker, Helm, Terraform, and cloud deploy configs, clone the starter kit directly.

Existing directory

If you want to create a new Pixeltable project in an existing directory, run the command without a project name:

uvx pixeltable-new

Learn more

• Pixeltable Documentation
• Starter Kit
• AI Coding Skill
• Discord

License

This project is licensed under the terms of the Apache 2.0 license.