Caddy
Caddy Technologies Inc.
As of 9 Oct 2026, PAEval has compiled 15 capability records for Caddy from 13 source URLs; 1 are independent observations and 12 come from the vendor's own documentation. Assistant Benchmark (updated 8 Oct 2026) gives it 7.3 on its 1–10 scale, the mean of the 9 of 15 dimensions it scored, ranking #12 of 12 scored products. Scores from different evaluators use different scales and are not combined. PAEval does not publish its own score or rank for it.
Has cross-app writing, website tasks, and proactive follow-up; not just a chat interface
- Independent1
- Vendor12
- Third-party report1
- Not established1
15 records · 13 source URLs · 1 external result(s)
Capability summary
Strongest status per domain; every underlying record is listed below. Availability does not establish task success.
Positioning & scope
Handles family schedules, travel, shopping, budgeting, and small business responses.
Regions, eligibility & access
Public beta, free to start; list site records Free public beta, no standard payment price list.
The listed sources were reviewed, but sufficient public documentation was not found.
Limitations: Having a NYC landing page does not mean it is only open in NYC; the official website does not have a complete country/language list.
Autonomy & background work
Temporary/cyclic reminder; new users default to the local 8 o'clock morning briefing, which will be sent only if it is useful. Important changes in Gmail/Calendar/Slack/Linear can trigger check-in.
Active messages can be stopped while retaining individually created reminders and routines; only selected connection sources are checked.
Execution environments
Officials explicitly handle store, website, and airline check-ins in their own browser; no word on the OS/persistence of that environment or whether it supports user-native control.
Calls, messages & identity
You can read, draft and send from Gmail, read and write with Google Calendar; your own clear schedule can be updated directly, external invitations/emails/orders need to be reviewed.
Connectors & tools
Gmail, Google Calendar, Slack, Linear, Todoist; and Drive/Sheets reimbursement workflow. Plaid only queries transactions but does not transfer money.
Memory & personalization
The official website says that you can learn communication, working methods, and preferences with use; the complete memory editing interface is not explained.
Channels & interfaces
iPhone iMessage, Android RCS, voice messages and group chats; you can ask Siri to send text messages to Caddy.
Research & deliverables
Official use-cases include receipt screenshots to Drive/Sheets, file search, meeting action items, and travel research; full PPT/video/code generation specifications are unknown.
Collaboration & parallelism
Can join group coordination; other iMessage conversations without Caddy participation will not be read.
Reliability & recovery
The travel page uses final confirmation from the booking platform as the booking criteria; browser login/store availability will limit execution. Idempotence and breakpoint recovery are not disclosed.
Permissions, security & data
The policy states that Customer Data is not used to train AI; some providers can configure zero retention, but not all. You can turn off sending LLM interactions to the evaluation service provider, and the turn off will take effect in the future.
External evaluations & observations
The test inbox/Notion/Calendar/Slack in September was unavailable; there was no outbound call channel; seats were remembered on 9/16 but dietary preferences were omitted. Cannot test unconnected when the connector is not present.
External measurements & hands-on results
Different methods, samples and dates. Not combined into one success rate.
Assistant Benchmark ↗
- Tested
- 2026-09-16
- Version
- v0.2
- Source score
- 7.3
- Scored dimensions
- 9
- Evidence labels
- test: 9; observed: 0; unlabelled: 0