Thesys provides frontend infrastructure for AI products through its C1 API and Crayon React SDK. Instead of returning plain text, C1 generates structured UI components like forms, tables, charts, and layouts directly from LLM responses. It works as an OpenAI-compatible endpoint, so you can drop it into existing AI agents with minimal code changes. The platform supports multiple LLMs including OpenAI and Anthropic, lets you bring your own React components, and handles streaming, error recovery, and fallback routing out of the box.

• Generative UI API (C1): Converts LLM responses into live, interactive UI components like forms, tables, and charts instead of plain text. Your AI agents can now respond with fully rendered interfaces.
• OpenAI-Compatible Endpoint: Works as a drop-in replacement for OpenAI API calls, so you can integrate it into existing projects with minimal code changes across any language or framework.
• Crayon React SDK: An open-source React UI library used with the C1 SDK to render generated interfaces, manage styling, and support custom themes.
• Custom Component Support: Bring your own React components with custom logic, styling, and interactions. Build specialized elements like weather cards or flight seat maps that go beyond built-in components.
• Multi-LLM Support: Works with OpenAI, Anthropic, and other major LLM providers. If one provider goes down, C1 automatically retries or routes to a fallback so your users stay unaffected.
• Real-Time UI Streaming: Components appear progressively as they are generated, creating a responsive experience without waiting for the full response to complete.
• Artifacts API: Generate slides, reports, and multi-page documents directly from your AI applications, billed separately from LLM token usage.
• Enterprise Security: Includes enterprise-grade infrastructure, compliance support, and private deployments.

• Significant Cost Reduction: Reduces frontend development costs by up to 80% by eliminating the need for manually coding every UI variation your AI agent might need.
• Fast Integration: OpenAI-compatible API format means you can swap it into existing AI agent codebases with minimal refactoring, getting to a working demo quickly.
• Built-In Resilience: Automatic LLM fallback routing and response validation means your app stays up even when individual providers have issues.
• Flexible Design System: Supports any design system and allows full component customization, so generated UIs match your brand rather than looking generic.
• Active Open-Source Ecosystem: The Crayon SDK is MIT-licensed with strong community examples covering LangChain, CrewAI, CopilotKit, and other popular AI frameworks.

• Developer-Focused Tool: Designed for engineers, not non-technical users. There is no visual builder or no-code interface for creating generative UIs.
• LLM Quality Dependency: The quality of generated UIs depends on the underlying LLM's capabilities. Complex or unusual layouts may not always render as expected.
• Limited Public Pricing Transparency: Detailed per-token costs are not clearly published on the website, making it harder to estimate costs before signing up.
• Initial Setup Complexity: New users report that integrating C1 with existing systems and configuring custom components takes some effort to get right.

• AI-Powered Analytics Dashboards: A data team building an internal analytics copilot can use Thesys to let users ask natural language questions like "show me monthly revenue trends" and receive interactive charts and tables instead of raw text. This eliminates the need to pre-build every possible dashboard view.
• Conversational E-Commerce Agents: An online retailer can build a shopping assistant that generates product comparison tables, interactive filters, and checkout forms directly in the chat interface. Users complete their entire purchase flow without leaving the conversation.
• Enterprise Copilots and Internal Tools: Companies building AI copilots for sales, procurement, or HR can use C1 to generate approval forms, forecasting dashboards, and workflow interfaces on the fly, adapting the UI to each user's specific request rather than maintaining dozens of static screens.
• Interactive Educational Platforms: EdTech teams can create AI tutors that respond with quizzes, interactive diagrams, and progress trackers instead of walls of text. Each student gets a UI tailored to their learning path and current topic.
• Healthcare Data Interfaces: Healthcare providers can build patient-facing or clinician-facing tools where AI generates dynamic data visualization dashboards, intake forms, and summary reports adapted to the specific clinical context.
• Rapid Prototyping for AI Startups: Early-stage teams can use Thesys to build functional AI product demos fast without investing heavily in frontend development, testing product-market fit before committing to custom UI engineering.