skip to content
gigarouter gigarouter
models / embeddings · coming soon

MiniCPM-Embedding

openbmb/MiniCPM-Embedding

published Sep 2024 · updated Jan 2025

MiniCPM-Embedding is a bilingual text embedding model that produces dense vector representations for Chinese and English, with strong cross-lingual retrieval capabilities.

est. price
~$0.008
/ 1M tokens · estimated, set at launch
API providers
0
downloads / mo
2.5K

specs

TaskText Embedding
ArchitectureBidirectional attention with Weighted Mean Pooling, based on MiniCPM-2B
Parameters2.4B
Embedding Dimension2304
Max Input Tokens512
LicenseApache-2.0 (code); MiniCPM Model License (weights, free for academic and commercial use after registration)

best for

FAQ

What is the embedding dimension and max input length?

The embedding dimension is 2304 and the maximum input token length is 512.

What input format does MiniCPM-Embedding expect?

It supports an optional query-side instruction in the format "Instruction: {{ instruction }} Query: {{ query }}", or instruction-free mode as "Query: {{ query }}". Documents are input directly.

What is the model size and how does it compare to other embedding models?

MiniCPM-Embedding has 2.4B parameters. It achieves 76.76 NDCG@10 on C-MTEB/Retrieval and 58.56 on BEIR, outperforming many larger models in cross-lingual tasks.

What are the license terms for using MiniCPM-Embedding commercially?

The code is Apache-2.0. The model weights require following the MiniCPM Model License; they are free for academic research and free for commercial use after filling out a registration questionnaire.

How can I call MiniCPM-Embedding via the gigarouter API?

Use the gigarouter OpenAI-compatible endpoint with your API key, sending your text as input to the embeddings endpoint.

not yet live

We're benchmarking and onboarding MiniCPM-Embedding as a hosted, OpenAI-compatible API. Sign in for free credit and be ready when it lands, or tell us you want it and we'll prioritize it.

related embeddings models

compare all →