Choose your plan
Pricing
Start free. Upgrade when you need more investigations.
Free
Free
Enough to judge whether this is for you.
5 investigations/month
- + Free re-judging of saved sessions in the app and REST API
- + Publish and share results
- + Model Context Protocol (MCP) and REST API access
Analyst
$20/mo
For work that goes to someone else.
40 investigations/month
- + Free re-judging of saved sessions in the app and REST API
- + Publish and share results
- + Model Context Protocol (MCP) and REST API access
Pro
$99.99/mo
For work that goes to a client, a court, or a committee.
400 investigations/month
- + Free re-judging of saved sessions in the app and REST API
- + Publish and share results
- + Model Context Protocol (MCP) and REST API access
How usage is counted
- MCP
- Model Context Protocol, an open standard that lets AI assistants such as Claude call outside software. Arbiter uses it to connect to your assistant.
- Investigation
- One run of the reasoning engine on a question, its sources, claims and evidence, producing a verdict (a judgement). A run that fails without a verdict is not counted. For a second verdict on the same investigation, see Re-judging.
- Re-judging
- A new verdict on a saved investigation that already has one, for example after you add or change evidence. It is free in the app, through the REST API endpoint POST /api/v1/sessions/{id}/judge, and through MCP. Only a first judgement on an investigation and each call to the one-shot endpoint POST /api/v1/investigations/analyze count as an investigation.