EmbeddingsCC-BY-NC-4.0

jina-code-embeddings-0.5b

A compact encoder for finding code from natural language, technical questions, and related code.

Use API Official model card Full research detailsSelect for comparison

Base Model Family

Qwen2.5-Coder-0.5B

Deployment

Hugging Face AWS SageMaker Microsoft Azure Google Cloud Air-gapped

Related models

jina-code-embeddings-1.5b

Academic Publications

NeurIPS 2025 August 31, 2025 Efficient Code Embeddings from Code Generation Models October 03, 2025 • 7 minutes read Jina Reranker v3: 0.6B Listwise Reranker for SOTA Multilingual Retrieval New 0.6B-parameter listwise reranker that considers the query and all candidate documents in a single context window.September 04, 2025 • 6 minutes read Jina Code Embeddings: SOTA Code Retrieval at 0.5B and 1.5B Code generation LLMs → code embeddings: 0.5B/1.5B models achieve SOTA performance across 25 code retrieval benchmarks.

Compare models 0 / 3

Select two or three models to compare their published specifications.

Released: 2025-09-01
Parameters: 494M
Context length: 32K tokens
Output dimensions: 896
Inputs: text · code
Languages: Multilingual
Outputs: vector
Availability: api · huggingface · aws · azure · gcp · airgapped

Choose a representation

The catalog lists 896 output dimensions. Supported dimensions and input types depend on the model; check the request schema before switching a production index.

Matryoshka dimensions: 64 / 128 / 256 / 512 / 896

embeddingsclassifier

Model weights and hosted API access

The license shown above applies to the published model assets. Calling a hosted API and downloading weights for commercial self-hosting are separate arrangements.

These weights are not listed under a permissive commercial license. Review the official license and obtain the required authorization before commercial self-hosting.

Read the official license

Catalog checked on September 7, 2026. This is an independently prepared Chatsax guide to Jina models.