AICredits logo
Integrations

Hermes Agent

Configure Hermes Agent to use AICredits with an OpenAI-compatible base URL, your AICredits API key, and any supported model.

Use this page with an AI assistant

Opens a new chat with this docs URL and the correct AICredits base URLs.

Hermes Agent can use AICredits when it supports an OpenAI-compatible provider or custom base URL. You keep the normal OpenAI request format, but requests are billed through your AICredits INR wallet.

Settings

Use these values in Hermes Agent's provider or model settings:

SettingValue
Provider typeOpenAI-compatible / custom OpenAI endpoint
Base URLhttps://api.aicredits.in/v1
API keysk-your-aicredits-key
Chat modelopenai/gpt-4o-mini, anthropic/claude-sonnet-4.5, or another model from Models

If Hermes Agent asks for an OpenAI key, paste your AICredits key instead. If it asks for a base URL, use the /v1 URL exactly.

Environment Variables

If Hermes Agent reads OpenAI-compatible environment variables, set:

export OPENAI_API_KEY="sk-your-aicredits-key"
export OPENAI_BASE_URL="https://api.aicredits.in/v1"
export OPENAI_MODEL="openai/gpt-4o-mini"

Restart Hermes Agent after changing environment variables.

Use caseModel
Low-cost agent loopsopenai/gpt-4o-mini
Coding and tool useanthropic/claude-sonnet-4.5
Fast general chatgoogle/gemini-2.0-flash-001
DeepSeek workflowsdeepseek/deepseek-v4-pro

Test Prompt

After setup, run a small prompt:

Reply with only: AICredits connected

Then check Dashboard → Usage. You should see the request, model ID, token usage, and INR cost.

Troubleshooting

401 Unauthorized — Use your AICredits key, not an OpenAI or Anthropic key.

404 Not Found — Confirm the base URL is https://api.aicredits.in/v1, not just https://api.aicredits.in.

Model not found — Pick an active model ID from Models. Some tools hide unknown model IDs in a dropdown; use the custom model field if available.

Images API confusion — Vision/image-output chat models are not always /v1/images/generations models. Use the Images API filter when you specifically need image generation endpoint support.

On this page