Skip to main content

Documentation Index

Fetch the complete documentation index at: https://mind-e.co/docs/llms.txt

Use this file to discover all available pages before exploring further.

​
General

Mind-e is a platform for creating AI agents that answer questions from your documents, websites, and custom content. You upload your content, create an agent, and deploy it as a chat widget on your website or connect it to social media platforms.
Mind-e supports models from OpenAI (GPT-4o, GPT-4.1, GPT-5), Google (Gemini), Anthropic (Claude), and xAI (Grok). You choose the model for each agent in its settings. GPT-4o is the default.
Your documents are stored in Supabase (PostgreSQL and object storage). Sensitive tokens (social media API keys) are encrypted before storage. Each agent has an isolated vector namespace — agents cannot access each other’s data.

​
Documents & Knowledge Base

Document processing typically takes 10–60 seconds. If a document is stuck for more than a few minutes:
  • Delete the document and re-upload it
  • Check that the file isn’t corrupted (try opening it locally)
  • For PDFs, ensure the document contains text (scanned image-only PDFs may fail)
  • Check that you haven’t exceeded your storage limit
PDF (.pdf), Word (.docx), text (.txt), and Markdown (.md). See Limits for size and page limits per format.
  • Check that the document status shows Embedded (not processing or error)
  • Try asking the question differently — longer, more specific questions perform better
  • Enable query expansion in agent settings to improve retrieval for short questions
  • Review the sources shown with each response to see which documents are being referenced
  • Add a text snippet with the specific answer if the agent consistently misses it
Yes, you can upload multiple documents. Each one is processed independently. All documents contribute to the agent’s knowledge base.

​
Chat & Widget

  • Ensure the script is placed before the closing </body> tag
  • Verify that window.mindeConfig is defined before the widget script loads
  • Check that the agent ID is correct
  • Open your browser’s developer console and look for errors
  • Make sure your site’s Content Security Policy allows scripts from widget.mind-e.co
Response time depends on:
  • The AI model selected (larger models take longer)
  • The size of the knowledge base (more content = longer search time)
  • Whether streaming is enabled (streaming shows results progressively)
If speed is a priority, use a mini-tier model like GPT-4o Mini.
Yes. You can customize colors (solid or gradient), fonts, avatars, welcome messages, widget position (bottom-left or bottom-right), and chat direction (LTR/RTL). See Chat Appearance.

​
Credits & Billing

When your credit balance reaches 0, AI-powered features stop working. Chat messages, text-to-speech, and scenario triggers will fail. Upload your plan to get more monthly credits, or wait for the next billing cycle when credits reset.
No. Credits reset at the start of each billing cycle. Unused credits do not carry over.
Currently, credits are only available through subscription plans. There is no option to purchase credits separately.
Go to your dashboard and check:
  • Credit summary for a breakdown by feature and agent
  • Credit history for individual transactions
  • Agent analytics for per-agent usage

​
Social Integrations

Social integrations require a paid plan (Starter or above). If you’re on the Free plan, upgrade to enable integrations.
  • Check that the integration status shows Connected in the Integration tab
  • Verify the bot token is correct (get it from @BotFather)
  • Try disconnecting and reconnecting the integration
  • Check that your agent has content in its knowledge base
All social media integrations are automatically disconnected when you downgrade to the Free plan. You’ll need to reconnect them if you upgrade again.

​
API

Go to your dashboard settings, find the API Keys section, and click Create API Key. See API Keys for details.
  • Check that your API key is correct and active
  • Ensure you’re using the Authorization: Bearer <key> header format
  • If the key is scoped to a specific agent, make sure you’re using the right agent ID
You’ve hit the rate limit. Check the X-RateLimit-Reset header in the response for when you can retry. Reduce your request frequency or contact support for higher limits.