Skip to content

Ask the HotKV docs

Answers come only from these docs and can be wrong; check the linked pages before you rely on them.

AMEM.SEM.RECALL

Last updated
AMEM.SEM.RECALL agent_id k value [value ...]
Group
Agent memory
Complexity
O(log N) where N is the number of facts
Arity
-4 (at least 4 words, counting the command name)
Flags
read-only, slow
Keys
first 1, last 1, step 1
HotKV Cloud
Performance and Scale
On-Premise
requires the agent-memory license feature

Agent memory gives AI agents four kinds of memory. Semantic memory holds an agent’s long-term facts, each with an embedding your application computes, and recalls them by cosine similarity to a query embedding. AMEM.SEM.RECALL returns up to k facts most similar to the query embedding, most similar first. Note that k comes before the vector. Expired facts are skipped, so you can get fewer than k results.

Case RESP2 RESP3
Always Array of results. Each result is a flat array: "key", the key, "content", the content, "similarity", the cosine similarity as a string (all bulk strings). Empty for an unknown agent or a vector of the wrong size. Same

Errors:

  • ERR invalid k if k is not a positive integer.
  • ERR invalid vector component if a vector value is not a finite number.
  • ERR empty embedding vector if no vector values are given.
  • ERR wrong number of arguments for 'amem.sem.recall' command if the argument count is wrong.
  • NOLICENSE agent-memory requires a HotKV license. Contact sales@hotkv.com if your license does not include the agent-memory feature.
127.0.0.1:6379> AMEM.SEM.RECALL notes 1 0.1 0.2 0.31
1) 1) "key"
2) "fact-1"
3) "content"
4) "Paris is the capital of France."
5) "similarity"
6) "0.99990195"

The exact similarity digits depend on the stored vectors.

HotKV-only command. Redis and Valkey have no equivalent.

AMEM.SEM.CREATE, AMEM.SEM.STORE, AMEM.SEM.FORGET, AMEM.SEM.LIST, AMEM.SEARCH, AI features guide