Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.nano-gpt.com/llms.txt

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

​
Overview

Use POST /api/v1/messages/count_tokens to estimate the number of input tokens for an Anthropic Messages-format request. This endpoint is useful for request planning, context budgeting, and cost control. The result is an estimate and may not exactly match a provider tokenizer.

​
Endpoint

POST https://nano-gpt.com/api/v1/messages/count_tokens

​
Authentication

Use either header:
  • Authorization: Bearer YOUR_API_KEY
  • x-api-key: YOUR_API_KEY

​
Request Body

{
  "model": "claude-3-5-sonnet",
  "system": "You are concise.",
  "messages": [
    {
      "role": "user",
      "content": "Hello"
    }
  ],
  "tools": [],
  "tool_choice": null
}

​
Supported Content

The request uses Anthropic message format and supports:
  • text blocks
  • image and document-like blocks
  • tool use blocks
  • tool result blocks
  • system
  • tools
  • tool_choice

​
Response

{
  "input_tokens": 123
}

​
Example

curl -X POST https://nano-gpt.com/api/v1/messages/count_tokens \
  -H "Authorization: Bearer $NANOGPT_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "claude-3-5-sonnet",
    "system": "You are concise.",
    "messages": [
      { "role": "user", "content": "Hello" }
    ]
  }'

​
Notes

  • This is an estimate for planning and cost control.
  • It may not exactly match a provider tokenizer.
  • For generation, use POST /api/v1/messages.