Pricing
Tokens resell at the provider’s list price - zero markup, always. We charge for verification, which is the part with the value.
Free
Free
Enough to prove the shape works on your own traffic.
- ✓1,000 verified requests / month
- ✓Schema + groundedness checks
- ✓7-day receipt history
- ✓One-line base URL swap, OpenAI- and Anthropic-compatible
- ✓Full MCP server and REST API access
llm11 Pro
$49/mo
The full verification ladder at production volume.
- ✓25,000 verified requests / month
- ✓All four verification rungs, including cross-model and self-consistency
- ✓Per-task-type policy engine
- ✓90-day receipt history
- ✓Unlimited team seats
llm11 Scale
$249/mo
Higher volume, longer history, and shadow replay before you switch models.
- ✓250,000 verified requests / month
- ✓Everything in Pro
- ✓1-year receipt history
- ✓Signed zero-retention attestation
- ✓Shadow replay (coming soon)
How tokens are billed
Every model call - the primary answer, an escalation, or a fallback LLM judge when Jev isn’t configured - is debited from a prepaid credit balance at the exact cost OpenRouter reports for that call, in real time, with no markup. Jev’s own triage and judgment calls cost nothing extra: its output tokens are unmetered. Top up any time from your dashboard; requests stop cleanly at a zero balance rather than running you into debt.
FAQ
What counts as a ‘verified request’?
One call through /v1/chat/completions or /v1/messages that runs at least the schema check - every request, since that rung always runs.
What happens if I reach the included volume?
Requests pause until the next calendar month. Token spend remains prepaid, so there are no surprise overage invoices.
Can I self-host?
Not in this build - see the roadmap in our docs.