Lumen products
One assistant, several front doors. Everything below is live today — each card links to the real surface, not a waitlist.
Ask & answer
Lumen Ask
Included in: Free · 50 answers a day, no card. Lite 300, Pro 600, Team 1500/seat, Max 3000, Power 10000 a day.
The web answer engine. Ask a question, get an answer with citations, pick the model or let the cost-router pick for you. Free tier, no card.
Open Lumen →Lumen Suite
Included in: Free · one Empire sign-in spans every module.
The unified, module-tabbed shell that puts Ask and the connected Empire modules behind one sign-in.
Open the Suite →Shareable answers
Included in: Free · public read-only links, no account needed to open one.
Every finished answer gets a public, read-only link at /s/<id> so you can hand a result to someone without an account.
Agent
Agent mode
Included in: Free gets a real taste (3 steps, 5 runs a day). Lite 6 steps / 40 runs, Pro 8 / 100, Team 10 / 250, Max/Power highest.
Switch Ask into a step-by-step agent: it plans, calls connectors, reads the result and decides the next step instead of answering in one shot.
Lumen Local Agent
Included in: Free · any signed-in account can pair one.
A small Python agent you run on your own machine. It gives Lumen scoped read (and optionally write) access to one directory, so the agent can work on your real files and run your build. Opt-in, paired per session, revocable.
Read the consent terms →Browser extension
Included in: Free · any signed-in account.
Runs the agent inside your own browser session, so it can act on pages you are already signed in to instead of a blank cloud browser.
Install →Connectors
Included in: Web search, web fetch, GitHub and calculations on Free and Lite. The browser, terminal and workflow connectors need Pro or above.
Google and GitHub OAuth, plus the shared Empire connector catalogue the agent can call as tools during a run.
Browse connectors →Lumen-Pro
Included in: A separate subscription (from $39/mo) — not part of a Lumen plan.
Scheduled, always-on agentic automation for teams: visual flows, saved connections and run monitoring. The unattended half of the product.
Open Lumen-Pro →Developers
Answer API
Included in: Free · every signed-in account gets a key, 100 requests a day. Metered pay-as-you-go lifts the daily cap.
POST /api/v1/answer — the same answer engine behind an API key, with the quick, deep and agent modes exposed as a parameter.
Lumen Router (LLM Gateway)
Included in: A separate product with its own keys and plans.
An OpenAI-compatible /v1 endpoint in front of many providers. Point an existing SDK at it, keep your code, and let the router pick the cheapest model that can answer the request.
Model roster
Included in: Free to read. Premium models run from Lite upwards, or on your own key (BYOK) on any tier, Free included.
The live list of every model the router can reach, which tier unlocks it, and which origins are excluded — served as JSON, always current.
View roster JSON →Lumen API guide
Included in: Free · public documentation.
The standalone quickstart for the cost-router: what it classifies, how it picks, and how to migrate an existing integration.
Quickstart →Everywhere else
Install Lumen as an app
Included in: Free · no account needed to install.
Lumen is an installable progressive web app on desktop and mobile — an offline shell, a home-screen icon and a standalone window, with no store download.
App manifest →Plans
Start free. Upgrade when you hit the daily limit, need the better models, or want the agent to run unattended. Our top tiers undercut the named competition: Max beats Claude Max (from $100/mo, claude.com/pricing); Power beats ChatGPT Pro, Perplexity Max and Google AI Ultra's top tier (all $200/mo).
See plans →Enterprise
Included in: Custom seat counts and contract — talk to sales.
Seat-based plans, a private model policy and an audit trail for regulated teams. Talk to us before you buy.
Contact →How the pieces fit
| If you want to | Use |
|---|---|
| Ask a question and get a cited answer | Lumen Ask |
| Have the assistant do multi-step work for you | Agent mode, inside Ask |
| Let the agent work on files on your own computer | Local Agent |
| Let the agent act in pages you are signed in to | Browser extension |
| Run agents on a schedule, unattended | Lumen-Pro |
| Call the answer engine from your own code | Answer API |
| Cut your existing LLM bill without changing your code | Lumen Router |