This project ships two TEI binaries under the tei/ folder, built from v1.7.0:
tei/tei-metal– for Apple Silicon (M1/M2) using the Metal backendtei/tei-onnx– for x86_64 using the ONNX Runtime backend
Feel free to build your own from source:
git clone https://github.com/huggingface/text-embeddings-inference.git
cd text-embeddings-inference
# On x86_64 with ONNX backend (recommended)
cargo install --path router -F ort
# On x86_64 with Intel MKL
cargo install --path router -F mkl
# On Apple Silicon (M1/M2) with Metal
cargo install --path router -F metalYou can also run the TEI router standalone:
# e.g. on CPU:
text-embeddings-router --model-id YOUR_MODEL_ID --port 8080Note: on Linux you may need OpenSSL & gcc:
sudo apt-get install libssl-dev gcc -y
cargo run --release -- \
-f your_dataset.surql \
-t pinecone \
--embedding-provider tei \
--tei-binary-path tei/tei-metal \
--embedding-model nomic-ai/nomic-embed-text-v2-moe \
--dimension 768--tei-binary-path : path to tei-metal or tei-onnx Leave --embedding-url empty to start a local server