Tokenize

Tokenize helper

Estimate token usage before sending a request.

Endpoint

POST /api/v1/tokenize

Provide the same high-level inputs you would send to chat endpoints. You can send either text (string) or messages (array), and optionally system and model.

curl -X POST "https://api.kushrouter.com/api/v1/tokenize" \
  -H "Authorization: Bearer $API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gpt-4o-mini",
    "messages": [
      { "role": "user", "content": "Draft a customer onboarding email." }
    ]
  }'

Example response:

{
  "success": true,
  "model": "gpt-4o-mini",
  "tokens": 56
}

Typical workflow

  1. Call /api/v1/tokenize with your intended prompt.
  2. Inspect tokens and adjust prompts or models as needed.
  3. Send the final request to /api/v1/messages or the OpenAI/Anthropic endpoints.

Notes

  • Supports the same validation rules as the live endpoints; invalid payloads return HTTP 400.
  • Works with OpenAI and Anthropic style messages, respecting provider-specific parameter enforcement.
  • Subject to rate limit; excessive calls may return HTTP 429.
  • This endpoint estimates tokens only and does not execute the request.

Further reading

Sdk ExamplesTroubleshooting