BGE-M3 on Gonka

Multilingual text embeddings for search and RAG.

BAAI

What is BGE-M3?

BGE-M3 is BAAI's open-source embedding model: it turns text into 1024-dimensional vectors for semantic search, retrieval-augmented generation, clustering, and recommendations. It covers 100+ languages and accepts inputs up to 8,192 tokens.

On Gonka Broker the model is hosted on our own infrastructure and served through the standard OpenAI-compatible /v1/embeddings endpoint; any OpenAI SDK or RAG stack works unchanged. Billing is per input token; embeddings have no output tokens.

What BGE-M3 is best for

RAG pipelines

Drop-in embeddings for LibreChat RAG, Open WebUI, n8n vector stores, LlamaIndex, and similar stacks.

Semantic search

Dense retrieval over documents with an 8,192-token input window.

Multilingual matching

Cross-lingual retrieval across 100+ languages: query in one language, match content in another.

Clustering and dedup

Normalized vectors, ready for cosine similarity out of the box.

Key specs

Max input: 8K tokens
Dimensions: 1024
Modalities: Text in · vectors out
Features: Batch inputs, Base64 encoding
Model ID: BAAI/bge-m3

Pricing in USD

$0.01 / 1M input tokens

Embeddings are billed on input tokens only; there are no output tokens. Your USD rate is locked in when you top up, and it won't change until you've used the balance. No crypto, wallets, or tokens.

Or try BGE-M3 free: the monthly free tier covers about 35M tokens of this model. No card required.

See full Gonka API pricing

How to use the BGE-M3 API

The BGE-M3 API on Gonka is OpenAI-compatible, served through the standard /v1/embeddings endpoint. Point any OpenAI client at https://proxy.gonkabroker.com/v1, use your gnk-prx- key, and set the model to BAAI/bge-m3.

from openai import OpenAI

client = OpenAI(
    base_url="https://proxy.gonkabroker.com/v1",
    api_key="gnk-prx-YOUR_KEY_HERE",
)

resp = client.embeddings.create(
    model="BAAI/bge-m3",
    input="Text to embed",
)
print(resp.data[0].embedding)

import OpenAI from "openai";

const client = new OpenAI({
  baseURL: "https://proxy.gonkabroker.com/v1",
  apiKey: "gnk-prx-YOUR_KEY_HERE",
});

const resp = await client.embeddings.create({
  model: "BAAI/bge-m3",
  input: "Text to embed",
});

console.log(resp.data[0].embedding);

curl https://proxy.gonkabroker.com/v1/embeddings \
  -H "Authorization: Bearer gnk-prx-YOUR_KEY_HERE" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "BAAI/bge-m3",
    "input": "Text to embed"
  }'

How to create a Gonka API key

Run BGE-M3 on Gonka

Get started

Already have an account? Sign in