“How easy it was to deploy. We basically created client portals where I could put the chatbot directly into that portal so that a client would be able to access it while maintaining security.”
Developer API
#1 RAG API for Developers
RAG API with RESTful endpoints, comprehensive SDKs, and enterprise-grade reliability.
pip install customgpt-clientfrom customgpt_client import CustomGPT
client = CustomGPT(api_key="your_api_key")
response = client.send_request(
prompt="How do I use the SDK?")
print(response)RAG deployment
Deploy Enterprise-grade RAG applications faster with fully-managed RAG API
-
RAG API built for developers
Easy to use APIs and SDKs get you started in minutes. Our developer cookbooks will help you ship in minutes instead of months.
-
#1 in benchmarks
Our serverless RAG API ranked #1 in independent benchmarks for both accuracy and hallucinations.
-
Unmatched integrations
Integrate into any system using our RAG API or leverage hundreds of pre-built integrations. PDFs, Google Drive, Notion, Confluence and more.
-
Easy deployment widgets
Deploy using our RAG API or embed using out-of-the-box chat UI widgets, leveraging out-of-the-box UI.
Why CustomGPT.ai
Built for business, without the complexity
Leading example
CustomGPT.ai Is recognized as an industry leader in anti-hallucination
Why CustomGPT.ai?
Integrate a RAG API from the #1 RAG-As-A-Service platform. Leverage your business data into any workflow you have, and get accurate, secure, and hallucination-free responses, tailored for your content and environment, all in 3 lines of code.
#1 in accuracy and anti-hallucination
Experience industry-leading accuracy, validated by rigorous third-party benchmarks, enhancing your products with reliable and effective AI capabilities.
Save time building
Benefit from continually optimized RAG algorithms developed over thousands of hours and state-of-the-art RAG pipeline without the complexity of building and maintaining your own RAG system.
Security & Reliability
Enterprise-grade security
Data Use & Responsibility
Customer stories
What our customers say
See how teams use CustomGPT.ai to build AI they can trust.
-
-
-
“We chose CustomGPT.ai because of its accuracy, scalability, and ease of use. The no-code features are unmatched in the market.”
-
“We wanted the opportunity to be able to easily add podcast episodes and other multimedia content. So that was something in CustomGPT.ai that stood out to us.”
Have Questions About CustomGPT.ai's RAG API?
Why can’t I just build this with LangChain?
You certainly can, but I would urge you to read these white papers we have authored and make a wise choice that is right for your business.
Why wouldn’t I just build a RAG system myself?
With CustomGPT.ai’s RAG API, get up and running in minutes, not months. Benefit from continually optimized retrieval algorithms and state-of-the-art RAG pipeline without the need for in-house AI expertise. Redirect your valuable engineering resources towards developing unique features that set your product apart in the market. Read more about the argument for CustomGPT.ai here.
What kind of data do I need to provide to use the CustomGPT.ai RAG API?
You can provide any data that is relevant to your business operations, including files and websites or sitemaps. This includes product information, customer service logs, FAQs, and any other data that will help the AI understand and respond to queries specific to your business.
How can I start testing the RAG API?
We provide a sandbox environment via our documentation hub that allows you to test your API calls in a safe and controlled setting before going live. This environment queries your active projects in the application in real time.
How frequently is the RAG API updated?
We regularly update the API to introduce new features, enhance functionality, and incorporate feedback from our users. This is what has made us #1 in both accuracy AND hallucinations (a rare feat!)
We ensure that changes are backward compatible to protect existing integrations.
How do I get started with the CustomGPT.ai RAG API?
To start using the CustomGPT.ai API, you first need to sign up for an account on our platform. Once registered, you will be able to generate an API key which you can use to authenticate your API requests. Our documentation provides detailed guidelines on how to make API calls, manage your data, and configure your chatbot settings. Find out how to get started.
What are the RAG API rate limits, and how are they managed?
Our API rate limits are designed to ensure optimal performance and service reliability for all users. The specific limits depend on your subscription plan. If you need higher limits, you can discuss upgrading your plan with our sales team. You can monitor your API usage through our dashboard, which provides real-time analytics.
How do I handle errors or unexpected responses from the RAG API?
Our API documentation includes a comprehensive list of potential errors, their meanings, and suggested actions to resolve them. If you encounter an unexpected response or an error that persists, please contact our technical support team for immediate assistance.
Can I customize the RAG API responses?
Yes, you can adjust the behavior and style of the chatbot’s replies via our innovative ‘Persona’. Additionally, you can build your own data pipeline via our API to integrate complex logic or additional data sources that influence the chatbot’s responses.
What data formats does the CustomGPT.ai RAG API support?
Our API supports thousands of file formats for data upload. In addition, unlike other RAG services, our API can ingest entire websites and web data via sitemaps. Responses are returned in JSON format as output. This format ensures compatibility with a wide range of systems and simplifies the integration process. Detailed examples of request and response structures are available in our API documentation.
Is your code open-source?
Our Python SDK is open-source, and we also have a separate cookbook repository showing in-depth examples of how to use the SDK. Our application and RAG system are proprietary and closed-source.
Is my data secure with CustomGPT.ai?
Yes. CustomGPT.ai is SOC-2 Type 2 certified. Your data is fully encrypted and files are stored only when you allow them to be. Files are encrypted in transit, and encrypted at rest. We use secure vendors like AWS and Stripe.
Moreover, your Agents are fully self-contained, which means they do not share ANY data with other Agents, even within your own account.
By default, your Agent is private, which means only authorized users can query. We understand the importance of your data security and privacy and are fully committed to ensure that the chatbot and your data are secure, confidential, and private. Read about our security standards.
What support is available for CustomGPT.ai users?
All plans include email support through our helpdesk. In addition, customers on the Enterprise plan are assigned a dedicated account manager. Our team can assist with integration, troubleshooting, and any questions you may have about optimizing the use of your chatbot.
How do I integrate CustomGPT.ai with my systems?
Integration is facilitated through our easy-to-use API, which can be connected to your existing systems without significant changes to your infrastructure. Our support team is available to guide you through the integration process step-by-step. Read more about our existing integrations.
How does CustomGPT.ai ensure the accuracy of its chatbots?
Our chatbots are continuously tested against industry benchmarks to ensure they provide the most accurate responses. In independent benchmarks, CustomGPT.ai scored #1 in both accuracy and hallucinations (a rare feat!) . We are constantly striving to maintain that leaderboard ranking via proprietary optimizations to our RAG pipeline.