ollama_embed_search

Embed search (ad-hoc). FLAGSHIP. Rank AD-HOC candidates by concept similarity to a query. Pass query + candidates: [{id, text}]; server embeds everything, computes cosine, returns ranked [{id, score, preview?}]. Use this when you have in-memory candidates to compare; for persistent recall over memory/canon/doctrine use ollama_corpus_search instead. Does NOT return raw vectors to you.

Parameters

Parameter	Type	Required	Default	Description
query	string	yes	—	The question or concept to search for.
candidates	object[]	yes	—	Candidates to rank. 1-256 items; text is what gets embedded, id is what gets ranked.
top_k	integer	no	—	Return only the top K candidates (default: all of them).
preview_chars	integer	no	—	Include this many chars of each candidate’s text in the result (default 0 = no preview).

Full input schema

The JSON Schema below is generated from the same zod schema the server validates against at the wire — it cannot drift from runtime behavior.

{
  "$schema": "https://json-schema.org/draft/2020-12/schema",
  "type": "object",
  "properties": {
    "query": {
      "type": "string",
      "minLength": 1,
      "description": "The question or concept to search for."
    },
    "candidates": {
      "minItems": 1,
      "maxItems": 256,
      "type": "array",
      "items": {
        "type": "object",
        "properties": {
          "id": {
            "type": "string",
            "minLength": 1,
            "description": "Caller-chosen identifier returned unchanged in the ranking."
          },
          "text": {
            "type": "string",
            "minLength": 1,
            "description": "Candidate text to compare against the query."
          }
        },
        "required": [
          "id",
          "text"
        ]
      },
      "description": "Candidates to rank. 1-256 items; text is what gets embedded, id is what gets ranked."
    },
    "top_k": {
      "description": "Return only the top K candidates (default: all of them).",
      "type": "integer",
      "minimum": 1,
      "maximum": 256
    },
    "preview_chars": {
      "description": "Include this many chars of each candidate's text in the result (default 0 = no preview).",
      "type": "integer",
      "minimum": 0,
      "maximum": 500
    }
  },
  "required": [
    "query",
    "candidates"
  ]
}

Envelope

Every tool returns the standard envelope — tier_used, model, tokens_in/tokens_out, elapsed_ms, backend provenance, warnings. See Envelope & tiers.

Source

Schema + handler: src/tools/embedSearch.ts · registration: src/index.ts

---

Back to the Tool Reference.