Overlay

An AI workspace that thinks, remembers, creates, and acts — all in one place. Chat with top models, build persistent knowledge, generate media, and run automations without switching tabs.

🔗 Try Overlay

Powered By

How It Works

1. Bring your context — Upload files, save memories, create notes, and organize work into projects.
2. Ask or Act — Ask questions with full knowledge access, or switch to Act mode for tool use and automation.
3. Create & execute — Generate images and videos, run browser tasks, connect apps, and schedule automations.

Features

Core Capabilities

• Multi-Model Chat — Access top models from OpenAI, Anthropic, Google, xAI, Groq, and OpenRouter in one workspace
• Ask & Act Modes — Switch between answer synthesis (Ask) and tool execution (Act) depending on what you need
• Persistent Memory — Save preferences, facts, and standing instructions that compound over time
• Knowledge Base — Upload files, create folders, and search across your personal knowledge with semantic retrieval
• Project Organization — Group chats, notes, files, and context by project for focused work
• Media Generation — Create images and videos without leaving the workspace
• Voice Input — Record and transcribe audio directly into notes or chat
• Browser Automation — Run interactive browser tasks for live web work
• External Integrations — Connect Gmail, Calendar, Notion, GitHub, and 100+ apps via Composio
• Automations — Schedule recurring AI workflows that run on your behalf

Workspace Areas

Area	Description
Chat	Multi-model conversations with context from memories, files, and projects
Notes	Rich notebook editor with markdown, slash commands, and project linking
Memories	Durable facts and preferences that shape future responses
Knowledge	File storage with semantic search across documents and folders
Projects	Scoped workspaces for organizing related chats, notes, and files
Outputs	Gallery of generated images and videos with metadata and downloads
Integrations	Connected apps and tools for external actions
Voice	Audio recording and transcription flows

Tools & Actions

Knowledge & Memory

• Memory CRUD — Save, update, delete, and search personal memories
• File search — Lexical and semantic search across uploaded documents
• Note management — Create, edit, and organize notebook entries
• Knowledge retrieval — Hybrid semantic + keyword search across all saved context

Content Generation

• Image generation — GPT Image, Grok Image, FLUX, Seedream models with aspect ratio control
• Video generation — Veo 3.1, Seedance, Grok Video, Kling, Wan models
• Image-to-video — Animate static images into motion clips
• Reference-to-video — Place characters into new video scenes
• Motion control — Transfer motion from reference video to character images
• Video editing — Transform existing videos with text prompts

Automation & Execution

• Browser sessions — AI-controlled browser for interactive web tasks
• Daytona sandbox — Run code and CLI tasks in persistent sandboxes
• Scheduled automations — Interval, daily, weekly, or monthly recurring workflows
• Skills — Reusable instruction templates for common tasks
• MCP servers — Connect external tool servers via Model Context Protocol

Integrations (via Composio)

• Gmail, Google Calendar, Google Sheets, Google Drive
• Notion, Asana, Slack, GitHub
• LinkedIn, X (Twitter), Outlook
• And 100+ more apps

Models Supported

Chat Models

Provider	Models
OpenAI	GPT-5.4, GPT-5.4 Mini, GPT-4.1
Anthropic	Claude Opus 4.7, Claude Sonnet 4.6, Claude Haiku 4.5
Google	Gemini 3.1 Pro, Gemini 3 Flash, Gemma 4 26B
xAI	Grok 4.20
DeepSeek	DeepSeek V4 Pro, V4 Flash
Moonshot	Kimi K2.6, Kimi K2 Thinking (free)
MiniMax	MiniMax M2.7
Qwen	Qwen 3.6 Plus
GLM	GLM 5.1
Groq	GPT OSS 120B
NVIDIA	DeepSeek V3.2 (free), Kimi K2 Thinking (free)
OpenRouter	Free Router (auto-selects free models)

Image Models

• GPT Image 1.5, Grok Image Pro, Grok Image
• FLUX 2 Max, FLUX Schnell
• Seedream 5.0 Lite, Seedream 4.5

Video Models

• Veo 3.1, Veo 3.1 Fast
• Seedance v1.5 Pro
• Grok Video
• Wan v2.6 (T2V, I2V, R2V)
• Kling v2.6 (T2V, I2V, Motion Control)

Built With

• Next.js 15 — React framework with App Router
• Tailwind CSS — Styling
• Vercel AI SDK — AI streaming, tool calling, and multi-provider support
• Convex — Backend functions, data, and realtime sync
• WorkOS — Enterprise authentication and SSO
• Stripe — Billing and subscriptions
• Composio — External app integrations
• TipTap — Rich text editing
• Daytona — Code execution sandboxes
• Browser Use — Browser automation
• OpenRouter — Unified model API
• Shadcn/UI — UI components

Repository Layout

.
├── convex/                  # Backend schema, queries, mutations, actions, HTTP routes
├── src/app/                 # Next.js pages, layouts, and API route handlers
├── src/features/            # Web feature containers and feature-local helpers
├── src/components/          # Shared UI primitives, layout, and providers
├── src/server/              # Server-only auth, billing, storage, AI, and route services
├── src/shared/              # Isomorphic contracts and client-safe helpers
├── packages/                # Cross-surface packages and typed API clients
├── scripts/                 # Sanity scripts and one-off checks
├── docs/                    # Product, setup, testing, marketing, and implementation docs
├── AGENTS.md                # Local agent workspace notes
├── LICENSE.md               # License terms
└── SECURITY.md              # Security policy and reporting guidance

Local Development

Prerequisites

• Node.js 20+
• npm
• Convex account (dev + prod deployments)
• WorkOS credentials (for auth)
• Stripe test credentials (for billing)

Quick Start

# Install dependencies
npm install

# Configure environment
cp .env.example .env.local
# Edit .env.local with your credentials

# Run development server
npm run dev

For complete environment setup, see docs/start/quickstart.mdx and docs/configure/environment.mdx.

Convex Workflow

Push backend changes to both environments:

npm run convex:push:all    # Push to both prod and dev

Available Commands

Command	Purpose
npm run dev	Run the Next.js development server
npm run build	Create a production build
npm run convex:push:prod	Push Convex changes to production
npm run convex:push:dev	Push Convex changes to dev
npm run convex:push:all	Push Convex changes to both

Security

• Session cookies are encrypted and signed
• Transfer tokens are short-lived and hashed
• Provider keys isolated behind PROVIDER_KEYS_SECRET
• Sensitive logs redacted in chat and billing flows
• See SECURITY.md for full security guidance

Contributing

• Do not commit real secrets or customer data
• Keep public docs on placeholders
• Treat NEXT_PUBLIC_* values as public
• Prefer backend logic in convex/ with web handlers in src/app/api/

License

Overlay uses a split license model:

• Core product surfaces are licensed under AGPL-3.0-or-later.
• Reusable SDKs, contracts, protocol packages, and shared UI packages are licensed under Apache-2.0.
• Overlay branding is governed by the trademark policy.

See LICENSE.md, docs/legal/licensing.mdx, and TRADEMARKS.md.

---

Overlay — One workspace, many models, real memory, live action.