configuring-embedder

GrepAI Embedder Configuration

View, change, or troubleshoot the embedding provider and model used by grepai. Handles cascading changes (dimensions, re-indexing, workspace propagation).

Model Reference

Model	Provider	Dims	Speed	Quality	Languages
nomic-embed-text	Ollama	768	Fast	Good	English
nomic-embed-text-v2-moe	Ollama	768	Fast	Better	100+ langs
bge-m3	Ollama	1024	Medium	Excellent	100+ langs
mxbai-embed-large	Ollama	1024	Medium	Better	English
all-minilm	Ollama	384	Very Fast	Basic	English
text-embedding-3-small	OpenAI	1536	Fast (API)	Good	Multi
text-embedding-3-large	OpenAI	3072	Fast (API)	Excellent	Multi

OpenAI pricing: text-embedding-3-small ~$0.02/1M tokens, text-embedding-3-large ~$0.13/1M tokens. Typical project (10k lines) costs ~$0.001.

Workflow

1. Detect Current Configuration

Check MCP registration (claude mcp list, .mcp.json, ~/.claude.json), local config (.grepai/config.yaml), and workspace config (grepai workspace list, ~/.grepai/workspace.yaml).

Mode detection:

• MCP has --workspace {NAME} → workspace mode, config in ~/.grepai/workspace.yaml
• .grepai/config.yaml exists, no workspace MCP → local mode
• Both exist → workspace for search, local for chunking/ignore

Display: mode, config path, provider, model, dimensions, endpoint.

2. Ask: What To Do

Ask via AskUserQuestion:

What would you like to do?

○ Change embedding model (keep same provider)
○ Change embedding provider (e.g. Ollama → OpenAI)
○ View current config (done — already displayed above)
○ Troubleshoot embedding issues

If "View current config" — stop here, already displayed.

3. Change Model or Provider

Show available models for current provider (use model reference table above), ask user to pick, check availability (Ollama: verify pulled, offer to pull), apply to correct config file (workspace: ~/.grepai/workspace.yaml, local: .grepai/config.yaml).

For provider switch: collect provider-specific settings (endpoint, API key, parallelism for OpenAI, dimension detection for LM Studio).

See references/provider-changes.md for full model selection flows, provider-specific setup commands, and config file templates.

4. Re-Index

Embeddings from different models are incompatible — index must be rebuilt. Warn user, ask to re-index now or later. Clear old index (GOB: remove .gob files, Qdrant: delete collection, PostgreSQL: truncate tables), then run grepai watch.

See references/reindex.md for backend-specific re-index commands.

5. Troubleshoot

If user chose troubleshoot: check provider connectivity, model availability, config consistency (model/endpoint not swapped, dimensions match reference), workspace vs local mismatch. Report with OK/FAIL/WARN indicators.

See references/troubleshooting.md for check commands and diagnostic steps.

6. Print Summary

Print before/after comparison (provider, model, dims), config path, mode, and re-index status. If re-index started, show monitor commands. If manual, show re-index commands.