Pricing that scales with
your AI, not your headcount.

Anyone on your team with read access to the Gheist web app. We don't count AI agents, view-only auditors, or external contractors against your teammate cap. If you have 30 full-time teammates and 15 contractors who occasionally collaborate, you're at 30.

Nothing breaks. You move to overage at $0.10/credit, billed at month end with a 14-day grace period. Your agents keep querying, your integrations keep syncing. If you want certainty on your bill, set a hard cap in the billing console — when you hit the cap, capture slows to essential events only until the next period.

Yes, for 60 days. After that they expire. This matches usage-based conventions (AWS, Heroku, Vercel) and prevents stockpiling. If your usage is genuinely bursty, annual commit smooths it out.

Per-seat prices the wrong thing. The value Gheist creates is captured knowledge × agents that can read it — which correlates with your AI adoption depth, not your headcount. Credits map revenue to value. As a side benefit, you can add engineers freely under each tier cap, which removes the usual adoption friction.

Yes, on Enterprise. We ship a Docker-based reference deployment plus Terraform modules for AWS/GCP. Your corpus, embeddings, and agent queries never leave your VPC. You bring your own LLM provider (Anthropic, OpenAI, or a private model). We provide dedicated engineer time during onboarding.

Starter through Business all include a 14-day free trial — full feature access, no credit card. After the trial you can convert, downgrade to Free, or cancel. We email you on days 7, 12, and 13 so you're never surprised.

Prepay 12 months up front, save 20% off list. Payment by card, ACH, or invoice. We'll generate a one-line order form for procurement if needed. You can still upgrade mid-year; we pro-rate the difference and extend your renewal date.

All 5 MCP tools (gheist_search, gheist_get, gheist_record, gheist_ask_human, gheist_log_outcome), the web app, provenance chains, and the full ingestion pipeline. The difference between tiers is credits, integration breadth, SLA, and enterprise features (SSO, audit logs, self-host).

Three triggers usually apply: (1) you need VPC / on-prem / BYO models, (2) you need SSO + SCIM + audit logs for compliance, (3) you're >500 engineers or credit consumption regularly exceeds Growth's 100k/month. Enterprise starts at $100k ACV and is priced on your specific seats + estimated usage.

Start on the Free tier (500 credits, no commitment). After 2-3 weeks you'll have real usage data and can pick a tier with confidence. Almost all teams land between Starter and Team for their first paid month.