V4 · Configuration

Three modes. One for every setup.

Choose a local core path, a local Ollama path, or a configured model-provider path. Modes A and B keep memory processing on infrastructure you control — a fit for EU data-residency — and a built-in EU AI Act self-assessment maps each mode's posture. Network behavior follows the complete configuration, not the mode label alone.

DEFAULT · RECOMMENDED
A

Local core path.
No LLM provider.

Mode A uses a local sentence-transformer for dense candidates without an LLM provider in the core memory-content path. Optional downloads, connectors, backups, proxies, and client applications retain their own network behavior.

  • EmbeddingLocal sentence-transformer
  • StorageSQLite-backed core + derived indexes
  • GPUNot required
  • Provider callNot required for the core path
  • Setup timeEnvironment-dependent
  • Published LoCoMo60.4% raw · 74.8% retrieval
Use when
  • Privacy-first orgs
  • CI/CD pipelines
  • Low-resource machines
  • Air-gapped environments
setup — Mode A
# Install SuperLocalMemory $ npm install -g superlocalmemory
# Run setup and select Mode A (default) $ slm setup
✓ Mode A active · local embedding path configured ✓ SQLite-backed core initialised ✓ Optimize cache remains opt-in
# Verify status $ slm status mode: A · core provider call: not required ✓ Review optional network features separately
LOCAL LLM
B

Local model path.
Ollama configured.

Mode B uses a configured Ollama endpoint for local-model embeddings and LLM-assisted behavior. Model acquisition and any non-local Ollama endpoint must be assessed as separate network paths.

  • EmbeddingOllama (local neural)
  • StorageSQLite-backed core + derived indexes
  • GPUDepends on the Ollama deployment
  • NetworkDepends on endpoint and model acquisition
  • Setup timeEnvironment-dependent
  • Published LoCoMoNo separate Mode B paper run
Use when
  • Existing local Ollama deployments
  • Teams that control their model endpoint
  • Teams already running Ollama
setup — Mode B
# Pull an embedding model via Ollama first $ ollama pull nomic-embed-text pulling nomic-embed-text:latest … done
# Switch the configured runtime to Mode B $ slm mode b
✓ Ollama detected · nomic-embed-text ready ✓ Local endpoint configured ✓ Optimize cache remains opt-in
# Test semantic recall $ slm recall "token expiry" ↳ JWT, refreshed every 15m · relevance 0.83 · stored confidence 0.99   answer confidence null · calibration uncalibrated
CLOUD
C

Provider-assisted.
Explicit data transit.

Mode C sends configured content to a model provider for assisted processing. Assess provider logging, retention, transfer, security, and legal terms for the complete deployment.

  • ProviderConfigured cloud LLM / embedding endpoint
  • StorageSQLite-backed core + derived indexes
  • GPUNot required
  • InternetProvider calls and optional integrations
  • Setup timeEnvironment-dependent
  • Published LoCoMo87.7% · Conv-30 / 81 questions
Use when
  • Teams that accept configured provider processing
  • Deployments with reviewed provider contracts
setup — Mode C
# Configure provider credentials using your secret manager $ export OPENAI_API_KEY=sk-...
# Switch to Mode C and select a provider $ slm mode c $ slm provider set openai
✓ Provider path configured ✓ Configured content may be sent to that provider ✓ Local core state remains operator-managed
# Inspect the complete configuration $ slm status ✓ Mode C active · review provider policy before use
Side by Side

Compare the three modes.

Every tradeoff, visible at a glance. No asterisks.

Feature Mode A Mode B Mode C
No internet required ✓ ✓ –
No GPU needed ✓ – ✓
No API key ✓ ✓ –
Setup time Environment-dependent Environment-dependent Environment-dependent
Published LoCoMo evidence 60.4% raw · 74.8% retrieval No separate paper run 87.7% · Conv-30 / 81
Embedding type Local sentence-transformer Ollama (neural, local) Configured provider path
Storage backend SQLite-backed core SQLite-backed core SQLite-backed core
Recommended for Privacy · CI/CD · Default Controlled Ollama endpoint Reviewed provider processing

Published V3 architecture evidence carried into V4: Mode A results cover 10 conversations and 1,276 questions; the 74.8% retrieval result discloses GPT-4.1-mini answer synthesis. Mode C covers Conv-30 only (81 questions) with cloud embeddings and GPT-4.1-mini. Mode B has no separately published LoCoMo run.

Governance by mode

Each mode, mapped to its EU AI Act posture.

SuperLocalMemory ships a per-mode EU AI Act self-assessment. It is a technical-control map, not a legal certification — applicability depends on your deployment, data, and operator role.

Mode A · Local core
Memory processing stays local with no generative AI in the core path. Assessed as meeting requirements — the strongest fit for EU data-residency.
Mode B · Local model
Local Ollama enrichment keeps processing on the machine with no external provider. Assessed as meeting requirements.
Mode C · Provider-assisted
Sends configured content to an external provider. Flagged non-compliant by the checker; assess provider terms, international transfers, and legal basis.

Regardless of mode: GDPR access / erasure / portability, a hash-chained audit trail, per-workspace isolation, and role-based access. See Governance & EU AI Act controls →

Inspectable portability

Change mode any time — your memories persist.

Modes share the local memory database, but provider and embedding changes can require materialization or re-indexing. Back up the configured data root, run the mode command, and verify status before resuming production traffic.

mode switch
# Switch from Mode A to Mode B $ slm mode --set B
✓ Mode changed: A → B · 528 facts preserved ✓ Re-indexing in background (Ollama neural)
# Check progress any time $ slm status mode: B · indexed: 528/528 · engine: ollama ✓ Re-index complete · all memories available
Common Questions

Questions about modes.