Pricing
14 days in Shadow Mode.
No card required.
Every workspace starts with a free agent spend audit: your agents run in Shadow Mode for 14 days, nothing is blocked, and the report shows what should have been stopped or sent to a person. Pick a plan when you want Enforce.
Developer
$49/mo
3 agents · 5K decisions/mo
- Shadow + Enforce
- Purpose-aware authority
- MCP + SDK
- Human review
- Tamper-evident audit trail
Team
$199/mo
25 agents · 50K decisions/mo
- Everything in Developer
- Webhooks
- Priority support
Prices in USD, billed monthly. Cancel any time.
Compare
What each plan includes
| Free audit | Developer | Team | |
|---|---|---|---|
| Price | Free, 14 days | $49 / month | $199 / month |
| Agents | 3 | 3 | 25 |
| Decisions / month | 5,000 | 5,000 | 50,000 |
| Shadow Mode and the Shadow report | ✓ | ✓ | ✓ |
| Purpose-aware authority | ✓ up to 5,000 checks | ✓ | ✓ |
| MCP server, TypeScript SDK, HTTP API | ✓ | ✓ | ✓ |
| Human review queue | ✓ | ✓ | ✓ |
| Tamper-evident audit trail | ✓ | ✓ | ✓ |
| Enforce | — | ✓ | ✓ |
| Webhooks | — | — | ✓ |
| Priority support | — | — | ✓ |
| API rate limit | 600 / min per key | 600 / min per key | 600 / min per key |
The free audit
How the 14 days work
- 01
Connect an agent
Sign up, describe what the agent is for and what it may spend, and add Neltava to Claude Code, Cursor or your code. It takes minutes.
- 02
14 days in Shadow Mode
Every spend is decided against your rules and its purpose, and recorded. Nothing is blocked. The clock starts with the agent's first real decision, not at sign-up.
- 03
Read the report, then decide
The Shadow report shows what would have been stopped or sent to a person. When you trust it, pick a plan and switch Enforce on, one agent at a time.
What the report measures, and when an agent is ready for Enforce: the Shadow report.
Do I need a card to start?
No. The 14-day audit is free and asks for no card. You only pay when you pick a plan.
When do the 14 days start?
With your agents' first real decision, not at sign-up — so a slow integration doesn't cost you days. Test decisions sent from the console don't start it.
What happens when the trial ends?
Nothing breaks. Shadow Mode never blocks, and every spend is still decided against your rules and recorded. Purpose is no longer assessed, and the report says so. Enforce needs a plan.
What counts as a decision?
Each spend your agent asks Neltava about. Retries with the same idempotency key return the original decision and aren't counted twice, and test decisions from the console aren't counted. Decisions are counted per calendar month (UTC).
What if I go over my decisions?
Decisions are never refused — a limit that stopped them would either break your agents or let spends through unchecked. Going over shows in the console, and we'll talk about the right plan.
What if I need more agents?
Agents are a hard limit: Developer has 3, Team has 25. Revoked agents don't count; paused ones do. Need more than 25? Write to hello@neltava.com.
Does Neltava hold or move my money?
No. Neltava decides; your agent pays the way it already does. There is no payment migration and no new card.
Can I cancel any time?
Yes, from the billing portal in the console. The plan runs to the end of the period, then the workspace returns to Shadow. Agents already in Enforce keep enforcing — protection is never switched off by billing.
Who handles billing and invoices?
Polar, our merchant of record, takes the payment and sends your invoices and receipts.
Do you train AI models on my data?
No. We don't sell your data, and we don't use your workspace or decision data to train AI models.