Cohere embedding readiness brief
Read-only Cohere embedding workflow. It lists embed-capable models, creates embeddings for explicit caller-provided text samples through recued-core.cohere.embed.create, and summarizes model/embedding readiness for search or RAG work. It never creates datasets or embed jobs, uploads files, calls streaming chat/generate endpoints, uses connectors, mutates fine-tuned models, or calls arbitrary Cohere routes.
How it works
Inspect the data fetches, transforms, gates, and output this recipe runs.
Process (7 steps)
models_raw
?
embed
?
models
default
Apply default
model_count
count
Count items in models
summarize
?
model_table
to_table
Format results as a data table
card
to_summary
Format results as a summary card
Settings
Configurable at install. Defaults shown — change them anytime in Recued.
focus
setting
=
Summarize whether the chosen Cohere embedding setup is ready for semantic search or RAG. Cover returned model options, embedding metadata, dimensions or output types visible in the response, warnings, and practical next steps.
texts
setting
=
cohere
setting
=
truncate
setting
=
END
input type
setting
=
search_document
max length
setting
=
900
embed model
setting
=
embed-v4.0
embedding types
setting
=
float
model page size
setting
=
20
output dimension
setting
=
1536
Trust & control
What installing this recipe would let it do. Recued grants these permissions at install — review them there before approving.