Ask every leading AI at once.
Nexus runs your prompt across Claude, ChatGPT and Gemini in parallel, shows you where they agree and diverge, and returns one synthesized answer with a verdict — so you decide with more than a single model's opinion.
One question. Every model. One answer.
Send one message to Claude, ChatGPT, Gemini — and up to two custom models — at the same time. See every answer side by side.
A synthesizer model distils the council into one answer: where they agree, where they diverge, and a plain verdict on which response was most useful.
Re-read any answer through CFO, COO, Legal, Risk and Customer lenses — plus blindspot, assumption and structural analysis. (Pro)
Bring your own provider keys. They are encrypted at rest and used only in-memory at query time — never exposed in the browser, never used to bill you.
Talk to the council and hear answers read back. Attach PDFs, spreadsheets, docs and images and ask across all of them.
Add any OpenAI-compatible model to the council and save system-prompt presets that prepend to every message. (Pro)
Four steps.
Add your own Claude, ChatGPT and Gemini keys once. Encrypted on arrival, never stored in plaintext.
One prompt goes to every model you enable. Watch each answer arrive in parallel.
Consensus, divergence and a verdict — so you leave with a decision, not five opinions.
Your conversations are encrypted at rest and synced to your account across web and mobile.
Plans.
You always pay AI providers directly with your own keys. Plans unlock how much you can do on top. Launching in 2026 — join the list and we'll let you in early.
- 5 queries per day
- Up to 3 saved chats
- Query up to 3 AIs at once
- Synthesis on every answer
- Voice playback & file attachments
- Encrypted, synced to your account
- Everything in Free
- Unlimited queries & chat history
- Up to 5 AIs in the council
- Blind taste test
- Cost & token tracking
- Expert analysis lenses
- System-prompt presets
- Custom AIs (up to 2 more)
- Synthesis pushback
- Everything in Pro
- Team workspace
- Shared prompt library
- Priority support
Good to know.
They are encrypted the moment they reach us and only decrypted in-memory to make a call on your behalf. They are never written in plaintext, never returned to the browser, and never used for billing.
You do, directly with each provider, using your own keys. Tumblehill never marks up or resells AI usage — Nexus is the layer on top.
Your conversations are encrypted at rest and synced to your account. We do not sell data or train on it. Each AI provider's own terms govern what you send to them.
The current flagships from Anthropic, OpenAI and Google out of the box, plus any OpenAI-compatible endpoint you add as a custom AI.