topic search

Topic search

Find a discussion even when it uses different words from your question. Search by meaning compares numerical representations of text, called embeddings, produced by a component you select. It needs stored transcripts and an installed local embedding model or an explicitly configured remote service. Use word search immediately if you have not configured one.

Choose where text is processed

A local embedding model processes text on this computer; it must already be installed. A remote endpoint receives transcript text when generating embeddings and your question when searching. Name that endpoint deliberately and choose its model identifier and vector dimensions. The tool does not download models. Previewing generation does not load a model or send transcript text.

The examples below use e5-small for a local model. Confirm it is installed before applying a batch; its name alone does not install it. Local and remote results need the same model/dimensions as their stored vectors to be comparable.

Search by meaning

Choose an already installed local model, or explicitly select a remote endpoint and dimensions:

zm search meetings "example plan" --model e5-small --max-chunks 1000 --json
zm search meetings "example plan" --model example-model --base-url http://127.0.0.1:8080/v1 --dims 384 --json

Before it reads, the search splits the current transcripts into passages again and embeds every passage that has no vector for that model yet, so a first search on a large store takes longer and, with a remote model, sends those passages to it. Continue with --cursor <next_cursor>; results are ranked per page only.

To prepare vectors ahead of time instead, preview and build them in batches:

zm meetings embeddings --model e5-small --json
zm meetings embeddings --model e5-small --yes --limit 100 --json

The first embedding command previews existing chunk status without opening a model, downloading files or generating vectors. --yes validates and opens the selected model before rebuilding chunks and embedding one bounded batch. Local models must already be installed; this command never downloads them. Remote endpoints are explicit; OpenAI endpoints use OPENAI_API_KEY. Remote models cannot be used with --offline.

Generating vectors

Generation uses current transcript revisions and rechecks their hashes before saving results. --after-transcript-id continues rebuilding chunks; --after-hash continues missing-vector work. These are separate cursors. Shared cached vectors remain available to other resources. A completed generation can emit its acknowledged receipt during a bounded cancellation grace period, then exit 130. Failed generation and cancelled queries emit no such receipt; a blocked output pipe can prevent delivery.

Searching

The command builds missing stored vectors first, then embeds the query and ranks the scanned candidates for the selected account and model. Continue with --cursor <next_cursor>; each page ranks its own candidates, rather than extending one global ranking. The report states scan coverage; index completeness and archive completeness remain unknown. --max-rows and --max-text-bytes bound transcript reads for generation and search. With MCP's server-configured by_meaning option, search reads existing vectors only; it cannot generate them.

Open a ranked passage with meeting context and read its surroundings: similarity is a candidate, not proof of a decision. If there are no candidates, check the selected profile/model, transcript availability and generation/scan coverage before concluding there is no relevant discussion.