# IronLabs Docs

## Docs

- [Foundry changelog](https://docs.ironlabs.ai/foundry/changelog.md): What's new in Foundry. Big drops, small wins, fixes — month by month.
- [Previous History](https://docs.ironlabs.ai/foundry/chat/history.md): How Foundry organizes, searches, and lets you reopen past chats.
- [Keyboard shortcuts](https://docs.ironlabs.ai/foundry/chat/keyboard-shortcuts.md): Every keyboard shortcut in Foundry, mapped to where it works.
- [Browser notifications](https://docs.ironlabs.ai/foundry/chat/notifications.md): Get a desktop ping when a long reply finishes, even if you switched tabs.
- [Voice input](https://docs.ironlabs.ai/foundry/chat/voice-input.md): Speak your prompts. Foundry transcribes and sends.
- [Adding a custom connector](https://docs.ironlabs.ai/foundry/connectors/adding-custom.md): Wire any API-key-auth service into Foundry so the assistant can call it.
- [Connectors](https://docs.ironlabs.ai/foundry/connectors/overview.md): Plug Foundry into external services — Exa, ElevenLabs, Firecrawl, and more — so the assistant can use real tools.
- [Pre-installed connectors](https://docs.ironlabs.ai/foundry/connectors/pre-installed.md): The eight services Foundry knows out-of-the-box. Paste the API key and you're done.
- [Zapier](https://docs.ironlabs.ai/foundry/connectors/zapier.md): Drop Foundry into any Zap and run prompts inside your existing automations.
- [Interface tour](https://docs.ironlabs.ai/foundry/get-started/interface-tour.md): Every important control in the Foundry chat UI, labeled.
- [What is Foundry?](https://docs.ironlabs.ai/foundry/get-started/introduction.md): IronLabs Foundry is the chat that picks the right model for every turn, and switches into Image, Video, or 3D mode when text alone won't cut it.
- [Quickstart](https://docs.ironlabs.ai/foundry/get-started/quickstart.md): Send your first chat on Foundry — under 2 minutes, no setup.
- [3D Website Mode](https://docs.ironlabs.ai/foundry/modes/3d-websites.md): Describe an experience, get a deployable 3D website. Currently in beta.
- [Image Mode](https://docs.ironlabs.ai/foundry/modes/image.md): Generate images inside chat with Nano Banana 2, Gemini 2.5 Flash, and friends.
- [Video Mode](https://docs.ironlabs.ai/foundry/modes/video.md): Brief, script, render, polish — generate cinematic video inside chat.
- [Memory](https://docs.ironlabs.ai/foundry/personalization/memory.md): Foundry remembering specifics — your projects, preferences, repeat context. Coming soon.
- [Personalization](https://docs.ironlabs.ai/foundry/personalization/overview.md): Teach Foundry how you like to be helped — tone, formatting, default context.
- [Reasoning effort](https://docs.ironlabs.ai/foundry/reasoning/effort-control.md): Instant, Low, Medium, High — the four levels of how hard reasoning models think.
- [Live reasoning view](https://docs.ironlabs.ai/foundry/reasoning/live-view.md): Watch the model think before it answers.
- [Reasoning models](https://docs.ironlabs.ai/foundry/reasoning/overview.md): How Foundry exposes reasoning — when it kicks in, what you see, how you control depth.
- [Importing Skills](https://docs.ironlabs.ai/foundry/skills/importing-skills.md): Drag-drop a ZIP, paste a GitHub URL, or pick from the official library.
- [Skills](https://docs.ironlabs.ai/foundry/skills/overview.md): Reusable instruction + tool bundles. Drop one in and the assistant gains a new ability.
- [Using Skills](https://docs.ironlabs.ai/foundry/skills/using-skills.md): Activate, stack, and manage Skills in the Skills manager.
- [Deep Research](https://docs.ironlabs.ai/foundry/sources/deep-research.md): Multi-step research — search, read, synthesize, cite. For when one query isn't enough.
- [Sources](https://docs.ironlabs.ai/foundry/sources/overview.md): Ground every reply in live data. Toggle Web, Deep Research, Reddit, X, Academic, or GitHub from the @ menu.
- [Reddit](https://docs.ironlabs.ai/foundry/sources/reddit.md): Ground replies in real-people Reddit threads — opinions, anecdotes, troubleshooting.
- [Web](https://docs.ironlabs.ai/foundry/sources/web.md): Ground replies in live web search results, with citations.
- [X (Twitter)](https://docs.ironlabs.ai/foundry/sources/x.md): Pull in posts from X for real-time news, tech announcements, and live sentiment.
- [Event-based tasks](https://docs.ironlabs.ai/foundry/tasks/event-based.md): Run prompts when something happens — not on a clock. Coming soon.
- [Tasks](https://docs.ironlabs.ai/foundry/tasks/overview.md): Run prompts automatically — on a schedule today, on events soon.
- [Scheduled tasks](https://docs.ironlabs.ai/foundry/tasks/scheduled.md): Run a prompt on a cron — hourly, daily, weekly, or anything in between.
- [Automate with a Scheduled Task](https://docs.ironlabs.ai/foundry/tutorials/automate-with-tasks.md): Build a daily brief that lands in your chat every morning at 8am.
- [Build a 3D website](https://docs.ironlabs.ai/foundry/tutorials/build-a-3d-site.md): Brief Foundry, iterate, deploy. A walkthrough of 3D Website Mode (currently beta).
- [Generate your first image](https://docs.ironlabs.ai/foundry/tutorials/first-image.md): A 5-minute walkthrough — open Image Mode, prompt, iterate, save.
- [Ground a reply with Sources](https://docs.ironlabs.ai/foundry/tutorials/ground-with-sources.md): Get a live, cited answer in 90 seconds.
- [Workflows · Blueprints](https://docs.ironlabs.ai/foundry/workflows/blueprints-coming-soon.md): Multi-step, reusable workflows. Coming soon.
- [Completions](https://docs.ironlabs.ai/refinery/api-reference/endpoint/completions.md): Send a list of messages and receive a model-generated completion. Supports streaming, web search, research mode, and external toolkits.
- [Model Select](https://docs.ironlabs.ai/refinery/api-reference/endpoint/model/select.md): Given an array of messages and the selection of LLMs you want to route between, `/model-select` will return a label for which LLM you should call.
- [Introduction](https://docs.ironlabs.ai/refinery/api-reference/introduction.md): IronLabs API Documentation
- [Refinery changelog](https://docs.ironlabs.ai/refinery/changelog.md): API, SDK, and docs releases for IronLabs Refinery.
- [AgentOpt](https://docs.ironlabs.ai/refinery/essentials/agentopt.md): Automatically optimize your agent's system prompt through closed-loop iterative benchmarking.
- [Client-Side Load Balancing](https://docs.ironlabs.ai/refinery/essentials/client_side_load_balancing.md): Learn how to configure client-side load balancing with IronlabsAI to distribute requests across multiple models for improved reliability and performance.
- [Custom Router](https://docs.ironlabs.ai/refinery/essentials/custom_router.md): Train and deploy your own AI model router to intelligently select the best model for each request.
- [Function Calling](https://docs.ironlabs.ai/refinery/essentials/function_calling.md): Learn how to use function calling with IronlabsAI to enable models to interact with external tools and APIs.
- [Images and Embeds](https://docs.ironlabs.ai/refinery/essentials/images.md): Add image, video, and other HTML elements
- [Markdown Syntax](https://docs.ironlabs.ai/refinery/essentials/markdown.md): Text, title, and styling in standard markdown
- [Model Select or Create](https://docs.ironlabs.ai/refinery/essentials/model-select.md)
- [Multimodal Inputs](https://docs.ironlabs.ai/refinery/essentials/multimodal_input.md): Learn how to use multimodal inputs with IronlabsAI to include images and PDFs in your messages.
- [Navigation](https://docs.ironlabs.ai/refinery/essentials/navigation.md): The navigation field in mint.json defines the pages that go in the navigation menu
- [GEPA vs MIPROv2](https://docs.ironlabs.ai/refinery/essentials/optimizer-comparison.md): Which prompt optimizer to use, why IronLabs runs both, and how the winner is picked.
- [Prompt Optimization](https://docs.ironlabs.ai/refinery/essentials/prompt_optimization.md): Automatically improve your prompts using advanced optimization techniques to achieve better results.
- [Reusable Snippets](https://docs.ironlabs.ai/refinery/essentials/reusable-snippets.md): Reusable, custom snippets to keep content in sync
- [Routing Lifecycle](https://docs.ironlabs.ai/refinery/essentials/routing-lifecycle.md): The five states every IronLabs request passes through, what triggers each transition, and what you can observe at each step.
- [Security & Isolation](https://docs.ironlabs.ai/refinery/essentials/sandbox-isolation.md): What is guaranteed to be separate inside IronLabs — and the mechanism that enforces each boundary.
- [Global Settings](https://docs.ironlabs.ai/refinery/essentials/settings.md): Mintlify gives you complete control over the look and feel of your documentation using the mint.json file
- [Structured Outputs](https://docs.ironlabs.ai/refinery/essentials/structure-output.md): Learn how to use structured outputs with IronlabsAI to get responses in a specific format.
- [Development](https://docs.ironlabs.ai/refinery/get-started/development.md): Learn how to preview changes locally
- [Overview](https://docs.ironlabs.ai/refinery/get-started/introduction.md): Ironlabs AI dynamically reduce your LLM Costs
- [Models supported](https://docs.ironlabs.ai/refinery/get-started/model.md): Supported LLM Provider & Models
- [Quickstart](https://docs.ironlabs.ai/refinery/get-started/quickstart.md): Make your first routed call in under 5 minutes.
- [Integrations](https://docs.ironlabs.ai/refinery/integrations/introduction.md): Ironlabs AI Integrations Documentation
- [Pricing](https://docs.ironlabs.ai/refinery/pricing/pricing.md): Choose the right plan to get the most out of Ironlabs AI

## OpenAPI Specs

- [openapi](https://docs.ironlabs.ai/refinery/api-reference/openapi.json)

## Optional

- [Dashboard](https://app.ironlabs.ai)
- [Open Chat](https://chat.ironlabs.ai)
- [Community](https://git.new/iai-discord)
- [Blog](https://ironlabs.ai/blog)