Pricing

Free

$0

Perfect for maintaining context graphs for your personal workspaces

  • ✓ 30 credits
  • ✓ Lightning-fast memory indexing and refresh; data stored locally
  • ✓ Unlimited automatic markdown structure setup
  • ✓ Unlimited retrieval requests
  • ✓ Bring your own key
  • ✓ Community support
Get Started

Pro

$19 /mo

For team and agent workspaces. Everything in Free, plus:

  • ✓ 200 credits
  • ✓ Hosted MCPs to connect to your agents and apps
  • ✓ Priority email support
Get Started

Custom

For products and supported deployments

  • ✓ Support real end-user workloads with the right deployment model
  • ✓ Get integration support, SLAs, and security review help
  • ✓ Scope credits, data residency, SSO, and on-prem needs together
Contact Us

Built from the suspicion that accumulated content is pointing somewhere — whether it's a personal vault or a product's user corpus.

Join the community

Frequently Asked Questions

Do I need an account? ▼

Yes for the current CLI flow. Login provisions the API key used by the CLI, workspace dashboard, and hosted memory features. Your vault files still stay on your machine unless you explicitly publish a hosted vault.

Is my data private? ▼

For local CLI use, Enzyme runs on your machine and your vault files stay on your device. Catalyst generation sends excerpts to the LLM provider you configure, or to Enzyme's default provider when you use the default setup.

For teams and products, deployment is scoped during onboarding: self-hosted by default, with managed infrastructure considered when it fits your data requirements.

See our privacy page for the full picture.

Why not just use Claude's large context window? ▼

Context windows work for one-shot exploration of a small corpus. Dump in 50 notes, ask a question, get an answer. They don't replace a persistent conceptual index for three reasons:

Attention is uneven. Long-context models are strong, but retrieval quality can still degrade when the relevant evidence sits deep inside a large prompt. Enzyme keeps the cross-cutting structure indexed instead of rediscovering it every session.

Cost scales with prompt size. For a vault of 2,000 notes, sending the corpus every session means paying for the same input repeatedly. Enzyme indexes once and keeps search-time queries local.

No persistence. Context windows reset every session. Enzyme's conceptual map persists and updates incrementally as you write. The agent doesn't rediscover your vault from scratch each time.

Where do credits fit? ▼

Credits are shown in your workspace dashboard. New accounts start with 30 credits. enzyme init uses 3 credits and enzyme refresh uses 1 credit when Enzyme provides the hosted LLM key. Bring your own provider when you want model calls billed directly through your own key.