Everything an assistant needs
to be trusted with real work
Context that persists
Projects remember your files, your tone, and the decisions you already made — across every conversation.
Inspectable reasoning
Expand any answer to read the chain of thought, the tools it called, and the exact passage each claim came from.
Deep Research
Mentr plans, searches, reads, and returns a fully cited brief — typically in under ten minutes.
Any model, one surface
Route between frontier, fast, and local models per task, with routing rules your team controls.
Shared prompt library
Version prompts like code. Publish to your team, fork a colleague's, and see which ones actually perform.
Private by default
Zero retention, regional processing, and SSO with SCIM provisioning on every enterprise workspace.
Six surfaces. One memory of your work.
Each surface is a real tool with its own history and controls — not a mode hidden behind a dropdown. They share the same projects, files, skills and permissions.
Long-context thinking over your files and connected sources, with the reasoning open to inspection.
An agent that reads the repo, writes the diff, runs the tests and opens the pull request.
Reference images, brand styles, inpainting and upscales — not a single prompt box.
Shot-by-shot generation on a real timeline, with camera moves and a render queue.
Package instructions, examples and tool scopes into something your whole team can run.
Every capability the model can reach, with the exact scope it was granted. Nothing silent.
The teams who moved first
"We replaced four tools with Mentr. The reasoning panel alone ended most of our review meetings."
"The first assistant our legal team trusts, because it shows where every sentence came from."
"Deep Research does a junior analyst's week in about nine minutes, with better citations."
Pay for capability,
not for seats you don't use
For individuals getting a feel for the surface.
- 40 messages a day
- Fast + balanced models
- 3 projects · 20 image credits
- 7-day history
For people who live inside the assistant all day.
- Unlimited messages
- Every frontier model + reasoning
- Deep Research, Vision, Voice, Code
- 500 image · 40s video credits / mo
Shared context, shared prompts, one invoice.
- Everything in Pro
- Shared projects, skills + pooled render credits
- SSO, SCIM, audit log
- Zero-retention processing
Answered
plainly
Still stuck? Talk to a human — we answer in under four hours on weekdays.
Every frontier model from the major labs, a fast tier for everyday drafting, and open-weight models we host in-region. You can switch mid-conversation and Mentr carries the full context across.