BAAI

bge-m3

bge-m3

View docs

Convert text into 1024-dimensional vectors for semantic search and retrieval.

Pricing: From 0 USD / 1M tokens
Output: Embeddings
Added to catalog: Oct 28, 2026

Input

Examples

Vector format

Default: "float"

Texts to embed*

Item 1

Item 2

Item 3

Default: Three semantic-search example texts

Estimate the cost first. Generation starts only after confirmation.

OutputFill example

3 inputs; 1024 dimensions per vector. Cosine similarity: related=0.7853, unrelated=0.4210. First eight coordinates: [0.002945489715784788,-0.007166733033955097,-0.05673351138830185,-0.025364981964230537,0.014633643440902233,-0.020261967554688454,0.0207122340798378,-0.024164272472262383]

README

Convert text into 1024-dimensional vectors for semantic search and retrieval.

Model capabilities

Input modalities: Text
Output modalities: Not confirmed

Parameters

Fields accepted in the request body for this model.

inputobject

Required

encoding_formatenum

Optional

Values: floatDefault: float

Pricing

USD / 1M tokens

Standard

Input0

Bill actual quantity

Minimum per call: 0 USD · Rounding is applied once to the total cost

Only listed usage types are supported for automatic billing

Pricing follows the selected specifications

Prices are billed per unit consumed with no monthly minimum. See billing guide for USD estimates and actual usage charges.

Code

View docs →

POST https://orux.top/api/v1/embeddings

Code examples

Request body

curl 'https://orux.top/api/v1/embeddings' \
  -H 'Authorization: Bearer YOUR_API_KEY' \
  -H 'Content-Type: application/json' \
  -d '{
  "input": [
    "社区咖啡馆提供安静的阅读空间。",
    "想找一家可以看书喝咖啡的店。",
    "明天的航班将在上午起飞。"
  ],
  "encoding_format": "float",
  "model": "bge-m3"
}'

Swipe horizontally to view the full content

Code examples

Response examples

{
  "object": "list",
  "data": [{ "object": "embedding", "index": 0, "embedding": [0.0023, -0.009, "..."] }],
  "model": "bge-m3",
  "usage": { "prompt_tokens": 8, "total_tokens": 8 }
}

Swipe horizontally to view the full content

Frequently asked questions

About bge-m3 pricing, capabilities and API access on Orux AI.

What is bge-m3?

bge-m3: Convert text into 1024-dimensional vectors for semantic search and retrieval.

How much does bge-m3 cost on Orux AI?

Input: 0 Credits; output: 0 Credits, per 1M tokens. See the pricing section on this page for bge-m3 pricing and applicable conditions. Charges depend on the selected parameters and recorded usage.

Where can I access the bge-m3 API?

Create an Orux AI API key and send an authenticated POST request to https://orux.top/api/v1/embeddings, using bge-m3 as the model identifier. Try the playground on this page and use the API example for the exact request format.