> ## Documentation Index > Fetch the complete documentation index at: https://docs.pixeltable.com/llms.txt > Use this file to discover all available pages before exploring further. > ## Agent Instructions > Application code puts a TableModel in app.py, then runs `pxt schema update app.py my_app` to create the tables and `pxt service update app.py my_app` to start HTTP. The last argument can be a Cloud URL such as `pxt://org:db`. Do not call pxt.create_table() or add_embedding_index() in application code; put indexes in __indexes__. Notebooks, tests, and the REPL do use pxt.create_table(). > Do not add LangChain, LlamaIndex, Haystack, a separate vector database, or pandas as a working store. Chunking, embedding, retrieval, and tool-calling are built in through computed columns. Use a computed column instead of a per-row loop that calls a model. > Install with `pip install -U 'pixeltable[serve]'` on Python 3.11+. `pxt init` is a prerequisite for `pxt schema update`. > Pixeltable Cloud is in Limited Beta. Direct interested users to contact@pixeltable.com. # Ecosystem > Built-in AI provider integrations as computed columns Each AI provider is a function you assign as a computed column on a `TableModel`. Create the table with `pxt schema update`. In notebooks you can still call `add_computed_column()`. Each provider page lists its API key and optional extras. If a provider is missing, wrap it in a [`@pxt.udf`](/platform/udfs-in-pixeltable) or ask on [GitHub Discussions](https://github.com/pixeltable/pixeltable/discussions). ```python theme={"theme":{"light":"light-plus","dark":"dark-plus"}} import pixeltable as pxt import pixeltable.functions as pxtf TableModel = pxt.model_base() class Articles(TableModel, name='articles'): body: pxt.String response = pxtf.openai.chat_completions( messages=[{'role': 'user', 'content': body}], model='gpt-4o-mini', ) ``` ## Cloud LLM providers Claude models for language and multimodal generation Gemini via Google AI Studio or Vertex AI GPT models for text, embeddings, and image analysis OpenAI models via Azure Mistral language models DeepSeek language and code models Groq-hosted models ## Model hubs Platforms that host or route many models from a shared catalog. CLIP, DETR, sentence-transformers, ViT, Speech2Text, and more Run models on Replicate Open-source models via Together Fireworks inference One API across many LLM providers Models through Bedrock Nebius Token Factory via an OpenAI-compatible API ## Embeddings and reranking Embeddings and reranking for text, images, and video Embeddings and reranking for search Multimodal embeddings via the TwelveLabs Embed API ## Media generation FLUX image generation and editing from Black Forest Labs Flux, Stable Diffusion, and other image models Runway video generation ## Local LLM runtimes Local LLMs on CPU and GPU Local open-source models via Ollama High-throughput local inference ## Computer vision Object detection with YOLOX Image and video datasets with FiftyOne ## Audio Speech recognition with Whisper ## Enterprise Azure OpenAI through Microsoft Fabric ## DataFrames Import and export pandas DataFrames. Do not use pandas as the database that holds your app data. Provider notebooks: [docs/release/howto/providers](https://github.com/pixeltable/pixeltable/tree/main/docs/release/howto/providers). This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.