Embedding Models - NanoGPT API Documentation

cURL

curl --request GET \
  --url https://nano-gpt.com/api/v1/embedding-models \
  --header 'Authorization: Bearer <token>'

{
  "object": "list",
  "data": [
    {
      "id": "text-embedding-3-small",
      "object": "model",
      "created": 1754480583,
      "owned_by": "openai",
      "name": "Text Embedding 3 Small",
      "description": "Most cost-effective OpenAI embedding model with dimension reduction support",
      "dimensions": 1536,
      "supports_dimensions": true,
      "max_tokens": 8191,
      "pricing": {
        "per_million_tokens": 0.02,
        "currency": "USD"
      }
    }
  ]
}

GET

embedding-models

cURL

curl --request GET \
  --url https://nano-gpt.com/api/v1/embedding-models \
  --header 'Authorization: Bearer <token>'

{
  "object": "list",
  "data": [
    {
      "id": "text-embedding-3-small",
      "object": "model",
      "created": 1754480583,
      "owned_by": "openai",
      "name": "Text Embedding 3 Small",
      "description": "Most cost-effective OpenAI embedding model with dimension reduction support",
      "dimensions": 1536,
      "supports_dimensions": true,
      "max_tokens": 8191,
      "pricing": {
        "per_million_tokens": 0.02,
        "currency": "USD"
      }
    }
  ]
}

Overview

The /api/v1/embedding-models endpoint provides a comprehensive list of available embedding models with detailed information including dimensions, max dimensions, token limits, pricing, and feature support. Use this endpoint instead of hardcoding embedding model capabilities. This endpoint is cacheable. Refresh it periodically because model availability can change.

Authentication

Authentication is optional but may enable user-specific features:

Header	Format	Required	Description
`Authorization`	`Bearer {api_key}`	Optional	API key for authenticated access
`x-api-key`	`{api_key}`	Optional	Alternative API key header

Response Format

Returns a list of all available embedding models with comprehensive details:

{
  "object": "list",
  "data": [
    {
      "id": "text-embedding-3-small",
      "object": "model",
      "created": 1754480583,
      "owned_by": "openai",
      "name": "Text Embedding 3 Small",
      "description": "Most cost-effective OpenAI embedding model with dimension reduction support",
      "dimensions": 1536,
      "max_dimensions": 1536,
      "supports_dimensions": true,
      "max_tokens": 8191,
      "pricing": {
        "per_million_tokens": 0.02,
        "currency": "USD"
      }
    },
    {
      "id": "text-embedding-3-large",
      "object": "model",
      "created": 1754480583,
      "owned_by": "openai",
      "name": "Text Embedding 3 Large",
      "description": "Highest performance OpenAI embedding model with dimension reduction support",
      "dimensions": 3072,
      "max_dimensions": 3072,
      "supports_dimensions": true,
      "max_tokens": 8191,
      "pricing": {
        "per_million_tokens": 0.13,
        "currency": "USD"
      }
    },
    {
      "id": "BAAI/bge-m3",
      "object": "model",
      "created": 1754480583,
      "owned_by": "baai",
      "name": "BGE M3",
      "description": "Multilingual embedding model with excellent performance across languages",
      "dimensions": 1024,
      "max_dimensions": 1024,
      "supports_dimensions": false,
      "max_tokens": 8192,
      "pricing": {
        "per_million_tokens": 0.01,
        "currency": "USD"
      }
    }
    // ... more models
  ]
}

Field Descriptions

Field	Type	Description
`id`	string	Unique model identifier to use in embedding requests
`object`	string	Always “model” for OpenAI compatibility
`created`	number	Unix timestamp of response creation
`owned_by`	string	Model provider (openai, baai, jina, etc.)
`name`	string	Human-readable model name
`description`	string	Detailed model description and use cases
`dimensions`	number	Default embedding vector dimensions
`max_dimensions`	number	Maximum embedding vector dimensions supported
`supports_dimensions`	boolean	Whether custom dimensions are supported
`max_tokens`	number	Maximum input tokens supported
`pricing`	object	Pricing information object

Pricing Object Structure

Field	Type	Description
`per_million_tokens`	number	Cost per million tokens in USD
`currency`	string	Always “USD”

Model Categories

OpenAI Models

High-quality embeddings with dimension reduction support:

text-embedding-3-small - Balance of cost and performance
text-embedding-3-large - Maximum accuracy
text-embedding-ada-002 - Legacy model

Multilingual Models

Support for multiple languages:

BAAI/bge-m3 - Excellent multilingual support
jina-clip-v1 - Multimodal CLIP embeddings

Language-Specific Models

Optimized for specific languages:

English: BAAI/bge-base-en-v1.5, BAAI/bge-large-en-v1.5, jina-embeddings-v2-base-en
Chinese: BAAI/bge-large-zh-v1.5, jina-embeddings-v2-base-zh, zhipu-embedding-2
German: jina-embeddings-v2-base-de
Spanish: jina-embeddings-v2-base-es

Specialized Models

Domain-specific embeddings:

jina-embeddings-v2-base-code - Optimized for code
Baichuan-Text-Embedding - General purpose
Qwen/Qwen3-Embedding-0.6B - Efficient with dimension reduction
Qwen/Qwen3-Embedding-4B - Higher quality Qwen embeddings
Qwen/Qwen3-Embedding-8B - Long-context Qwen embeddings
BAAI/bge-reranker-large - Reranking model
jina-embeddings-v3 / jina-embeddings-v4 - Newer Jina embedding models
gemini-embedding-001 - Gemini embeddings
doubao-embedding-large-text-240915 - High-dimensional embeddings (4096 dims)

Usage Examples

Basic Request

curl "https://nano-gpt.com/api/v1/embedding-models"

With Authentication

curl "https://nano-gpt.com/api/v1/embedding-models" \
  -H "Authorization: Bearer your_api_key_here"

Python Example

import requests

# Discover available embedding models
response = requests.get("https://nano-gpt.com/api/v1/embedding-models")
models = response.json()

# Display models sorted by price
for model in sorted(models["data"], key=lambda x: x["pricing"]["per_million_tokens"]):
    print(f"{model['id']}: ${model['pricing']['per_million_tokens']}/1M tokens - {model['dimensions']} dims")

JavaScript Example

// Discover available embedding models
const response = await fetch("https://nano-gpt.com/api/v1/embedding-models");
const models = await response.json();

// Find models that support dimension reduction
const flexibleModels = models.data.filter(m => m.supports_dimensions);
console.log("Models with dimension reduction:", flexibleModels.map(m => m.id));

Model Selection Guide

Use Case	Recommended Models	Rationale
General English text	`text-embedding-3-small`	Best price/performance ratio
Maximum accuracy	`text-embedding-3-large`	Highest quality embeddings
Multilingual content	`BAAI/bge-m3`	Excellent cross-language performance
Code embeddings	`jina-embeddings-v2-base-code`	Specialized for programming languages
Budget-conscious	`BAAI/bge-large-en-v1.5`	$0.01/1M tokens
Chinese content	`BAAI/bge-large-zh-v1.5`	Optimized for Chinese
Fast similarity search	Models with `supports_dimensions: true`	Can reduce dimensions for speed

Authorizations

Authorization

string

header

required

Bearer authentication header of the form Bearer <token>, where <token> is your auth token.

Response

List of available embedding models

object

string

required

Always 'list'

Example:

"list"

data

object[]

required

Array of embedding model objects

Show child attributes

Embeddings Image Generation (OpenAI-Compatible)

​Overview

​Authentication

​Response Format

​Field Descriptions

​Pricing Object Structure

​Model Categories

​OpenAI Models

​Multilingual Models

​Language-Specific Models

​Specialized Models

​Usage Examples

​Basic Request

​With Authentication

​Python Example

​JavaScript Example

​Model Selection Guide

Authorizations

Response

Overview

Authentication

Response Format

Field Descriptions

Pricing Object Structure

Model Categories

OpenAI Models

Multilingual Models

Language-Specific Models

Specialized Models

Usage Examples

Basic Request

With Authentication

Python Example

JavaScript Example

Model Selection Guide