Model Detail

linkup-sparseup-embed-v1 is a code generation model with 94M parameters released by Linkup-Platform. The model is registered under the feature-extraction pipeline tag on Hugging Face, distributed under the permissive apache-2.0 license.

linkup-sparseup-embed-v1 ships with 94M parameters. The apache-2.0 license is permissive, allowing commercial deployment and derivative work without per-seat fees, though attribution requirements still apply.

linkup-sparseup-embed-v1 is best fit for code completion, repository-scale Q&A, and pair-programming integrations, and semantic search, retrieval, and clustering pipelines. It is a less obvious choice for one-shot generation of security-critical code without review. Treat this as a starting matrix rather than a benchmark verdict — the right deployment usually depends on the specific evaluation suite that mirrors your workload.