ChatGPT
AI conversational assistant for answering questions, writing, and coding help
Morphic
AI-powered search engine with generative UI and streaming
Open-source AI search engine combining natural language understanding with generative UI. Supports multiple AI providers, search engines, and real-time streaming responses.
Overview
Modern AI Search with Generative UI
Morphic reimagines search by combining AI-powered natural language understanding with a dynamic generative user interface. Built on Next.js and the Vercel AI SDK, it delivers streaming responses that adapt in real-time, transforming traditional search results into interactive, contextual experiences.
Flexible Architecture for Developers
Designed for extensibility, Morphic supports multiple AI providers—including OpenAI, Anthropic, Google AI, Ollama, and Groq—alongside search backends like Tavily, SearXNG, and Exa. Developers can switch models directly from the UI, configure search depth and filtering, and deploy via Vercel, Docker, or Cloudflare Pages. Optional features include Supabase authentication, Redis-backed chat history, and reasoning models that expose their thought process.
Built for Self-Hosting and Customization
Whether you're prototyping a conversational search interface or deploying a production-ready alternative to traditional search engines, Morphic provides the stack: TypeScript for type safety, Tailwind CSS and shadcn/ui for rapid UI development, and Docker images for containerized deployments. The project is Apache 2.0 licensed and actively maintained, with comprehensive configuration options for video search, safe search filters, and custom time ranges.
Highlights
Generative UI with real-time streaming responses powered by Vercel AI SDK
Multi-provider support: OpenAI, Anthropic, Google AI, Ollama, Groq, and more
Flexible search backends including Tavily, SearXNG, and Exa neural search
Optional Supabase authentication, Redis chat history, and Docker deployment
Pros
- Extensive AI provider and search engine flexibility with runtime model switching
- Modern Next.js stack with React Server Components and TypeScript
- Self-hostable with Docker images and comprehensive deployment options
- Active development with Apache 2.0 license and strong community engagement
Considerations
- Requires API keys for AI providers and search services to function
- Optional features like chat history and video search add configuration complexity
- Generative UI approach may require learning curve for traditional search UX expectations
- Redis or Upstash needed for chat history persistence
Managed products teams compare with
When teams consider Morphic, these hosted platforms usually appear on the same shortlist.
Looking for a hosted option? These are the services engineering teams benchmark against before choosing open source.
Fit guide
Great for
- Developers building conversational search interfaces or AI-powered research tools
- Teams wanting self-hosted search with control over AI models and data
- Projects requiring flexible integration of multiple AI and search providers
- Organizations prototyping generative UI patterns for information retrieval
Not ideal when
- Users seeking a zero-configuration, API-key-free search solution
- Projects requiring traditional static search result layouts without streaming
- Teams without infrastructure for Redis or external AI provider management
- Use cases demanding sub-100ms response times for high-frequency queries
How teams use it
Internal Knowledge Search
Deploy a self-hosted AI search engine for company documentation with Supabase auth, enabling employees to query internal resources using natural language with reasoning transparency.
Research Assistant Interface
Build a conversational research tool that streams answers from multiple search providers, allowing users to switch between AI models and adjust search depth for academic or technical queries.
Custom Browser Search Engine
Integrate Morphic as a default browser search engine, replacing traditional search with AI-generated, context-aware results that adapt to user intent in real-time.
Multi-Model AI Experimentation
Prototype and compare responses across OpenAI, Anthropic, Ollama, and Groq models within a single interface, evaluating performance for specific domain queries without rebuilding infrastructure.
Tech snapshot
TypeScript99%
CSS1%
JavaScript1%
Dockerfile1%
Frequently asked questions
What AI providers does Morphic support?
Morphic supports OpenAI (default), Google Generative AI, Azure OpenAI, Anthropic, Ollama, Groq, DeepSeek, Fireworks, xAI (Grok), and any OpenAI-compatible provider. Models are configured in models.json with corresponding API keys.
Can I use Morphic without external search APIs?
Yes, you can self-host SearXNG as a search backend, eliminating dependency on commercial search APIs like Tavily. However, you still need an AI provider API key (or use Ollama locally) for language model functionality.
How does chat history work?
Chat history is optional and requires Redis (local or Upstash) for persistence. When enabled with Supabase authentication, users can save, retrieve, and share previous search sessions.
What deployment options are available?
Morphic supports Vercel one-click deployment, Cloudflare Pages, Docker Compose, and prebuilt Docker images from GitHub Container Registry. All methods require environment variables for API keys and optional services.
Is the generative UI customizable?
Yes, the UI is built with Next.js, Tailwind CSS, and shadcn/ui components. Developers can modify React Server Components, adjust streaming behavior via Vercel AI SDK, and override model configurations in models.json.
Project at a glance
Active- Stars
- 8,499
- Watchers
- 8,499
- Forks
- 2,321
LicenseApache-2.0
Repo age1 year old
Last commitlast month
Self-hostingSupported
Primary languageTypeScript
Last synced 2 days ago