Organized around the four modes, with model management, knowledge, artifacts, and team features grouped by category. Pick what interests you and go deeper.
modes
cross-mode handover paths
debate algorithms
connected AI services
Start from any of them.
One-on-one conversation. Actual PFC cost shown with every response.
Send one question to several models at once and view the answers side by side.
For/against/critic/adjudicator roles, plus a Red Team rebuttal pass.
Run a repeatable process as a template, up to 8 steps.
Co-edit the same draft with AI.
Cross modes or models without breaking the thread.
9 handover paths across Chat, Compare, Debate, and Workflow, all context-preserving.
39 AI services managed in one place — OpenAI, Anthropic, Google, xAI, Mistral, and more.
An independent review by a different model, available in every mode — Chat, Compare, Debate, and Workflow.
The foundation every mode is built on.
Ground answers in documents you upload.
AI output is auto-detected and saved, with cross-search and a trash bin.
Input and output templates managed separately.
Build your own personas and publish them in stages, with an approval flow.
State-managed conversations you can branch to try a different angle.
Provider-native search, switched on with a single toggle.
Dictate with your microphone and send straight away.
Actual cost shown per response, with a balance guard.
An iOS app (in TestFlight) with the core modes available.
Based on each vendor's public information as of July 2026.
| 項目 | Multi-AI aggregators | Single provider (ChatGPT / Claude) | Prompt Flow Studio |
|---|---|---|---|
| Model choice | Several vendors | One vendor's models only | Several vendors (OpenAI / Anthropic / Google / xAI / Mistral, and more) |
| Debate feature | Side-by-side comparison only | None | 3 algorithms + Red Team rebuttal + 7-metric scoring |
| Context-preserving handover | None | Within one vendor's models only | Across models and modes alike (9 paths) |
| Your documents | Capped by file count (e.g. 40–50 files) | Capped by file count per project | Measured in GB; file count isn't the limiting factor |
| Cross-AI audit log | None | Impossible by design | Available via API, export, and an in-app viewer |
| Cost visibility | Opaque, depends on plan | Metered on the API side | Actual cost shown per response |
See the full comparison on the compare page.
What you need to roll this out across an organization.
SAML SSO plus SCIM automatic provisioning.
Recording and lookup via API, export, and an in-app viewer.
Team sharing of threads and personas, with usage visibility.
Detects, masks, and restores PII. Team gets detection alerts, and Business gets automatic mask-and-restore.
The Free plan gets you the four core modes. Team features like SSO, SCIM, audit logs, and Security Filter are available from the Team plan and above. See the full picture on the business page.
Start with the free plan and try the four core modes.
Start free