OpenRouter-compatiblePublic docsNo Keypod rate limits

API Documentation

Keypod proxies OpenRouter-compatible requests with per-user usage tracking, project keys, and billing markup. Start with the setup steps, then copy the example that matches your stack.

Search OpenRouter models Create project keys

Get Started

Create a project, create a user key for that project, then send any OpenRouter model through Keypod's inference endpoint.

1. Create a project

Create a project in Keypod to group usage, billing, and user access.

2. Generate project and user keys

Combine them as project_key:user_key and send that as your bearer token.

3. Search OpenRouter for a model

Copy the model slug from OpenRouter and paste it into the modelfield in your request.

Color guide in the snippets

Project key + user keyModelURL

Notes

• Keypod does not impose platform rate limits.
• Use https://api.keypod.eyelinetech.com/v1 as your base URL.
• The main inference endpoint is POST /v1/chat/completions.

Examples

Pick the example for your stack. Each snippet uses the same project, key, and model placeholders.

cURL example

Use the inference endpoint directly with a combined bearer token.

bash

curl https://api.keypod.eyelinetech.com/v1/chat/completions \  -H "Authorization: Bearer kp_proj_xxxxxxxx:user_xxxxxxxx" \  -H "Content-Type: application/json" \  -d '{    "model": "openrouter/model-from-search",    "messages": [      {"role": "user", "content": "Hello!"}    ]  }'

Find Models

Search OpenRouter for the exact model slug you want, then paste it into your request body.

Open OpenRouter

Use the public model directory and search for the provider and model you want.

Copy the slug

Copy the exact model slug from OpenRouter so it can be passed through unchanged.

Paste into Keypod

Paste that slug into the model field in any request example.

Search OpenRouter models

Pricing

Keypod passes through upstream usage cost and adds platform markup.

Keypod applies a configurable 10% billing markup on top of the underlying OpenRouter response cost.

Billing is tracked per user, rolled up to the project level, and surfaced in the dashboard billing views automatically.

Cached pricing reference data is refreshed regularly and used when the upstream cost payload is unavailable.

Base URL

Use the custom domain for all client requests. The inference endpoint is shown below.

https://api.keypod.eyelinetech.com/v1/chat/completions