Skip to main content
Amazon Bedrock
Amazon Knowledge Base
Anthropic
Azure OpenAI
Cohere
Google Gemini
OpenAI
Custom AI Provider
Connect Retool to your choice of AI providers or custom LLMs to create powerful AI-enabled apps. Whether you're using third-party services like Mistral AI or your own models, you can build custom tools while maintaining security and compliance through your own infrastructure.
FAQs
Get answers to common Custom AI Provider questions
Retool's custom AI provider integration allows teams to connect any OpenAI-compatible endpointâincluding self-hosted models or private deploymentsâusing secure credentials stored as encrypted Retool resources.
Teams explicitly define which queries, fields, and inputs are included in prompts sent to custom providers inside Retool. Retool never sends data implicitlyâonly the data you configure is transmitted.
Yes. Retool's role-based access controls allow admins to limit which users or groups can create, edit, or run queries against custom AI provider resources.
Credentials for custom AI providers are stored as encrypted Retool resources and are never exposed to end users. Access is governed through Retool's permission model.
Yes. Retool supports environment-specific configurations, allowing teams to point different environments at different model endpoints or restrict custom AI usage per environment.
Retool's custom AI provider integration supports any endpoint that follows the OpenAI API specification, enabling teams to connect commercial, open-source, or internally hosted models.
Try Retool
Build with your favorite tools
More integrations
Explore more AI integrations
Amazon BedrockAmazon Knowledge BaseAnthropicAzure OpenAICohereGoogle GeminiOpenAI