Context Compression for LLMs in Rust | Signum

Signum is a Rust library and CLI for deterministic context compression before an LLM call. Publishing in the third release round.

Choose the audience that matches your question

The page contains three selectable readings of the same subject.

For consumers

Signum is a local context compressor. It makes a long note shorter before it reaches an AI system, follows the same rules for the same input and never sends your text elsewhere to shorten it. It publishes in the third round of Dweve's release programme, and the documentation arrives at docs.dweve.com that same day.

For businesses

Signum is a Rust library for prompt compression during LLM inference. It removes structurally redundant context before a paid model call; your text stays in your process, and the same input follows the same rule-based path. Signum publishes in the third round of Dweve's foundation release programme, with its documentation at docs.dweve.com the same day.

For engineers

Signum is a Rust library for deterministic prompt compression before an LLM call. It scores and shortens context inside the process that already holds it, returns the same result for the same input, and makes no model call or network request. The rules that decide what gets cut are worth reading: Signum publishes in the third round of the foundation release programme.