Start free. Keep what you use.

Drop in a spreadsheet and get your first answer in minutes. No card, no setup, no watches to configure. Upgrade the day you decide you want the answer to keep running.

Free
$0
Free to try

"Just answer my question"

Compute Ephemeral
Tokens One real question
Storage
  • Upload any spreadsheet, CSV, or data export
  • Full agent, not a demo
  • Enough tokens to finish one real question and follow-ups
  • Shareable link to the answer
  • Data not retained after session
Try free
Plus
$20
per month

"Keep my answers"

Compute 2 vCPU · 2 GB
Tokens 3× Starter
Storage 5 GB
  • Everything in Free
  • Persistent workspace
  • Connect Postgres, MySQL, REST APIs
  • Save Reports & Scripts
  • Weekly scheduling
  • Slack & email delivery
Start Plus
Max
$200
per month

"Run our data stack"

Compute 4 vCPU · 8 GB
Tokens 40× Starter
Storage 50 GB
  • Everything in Pro
  • Multi-site Federation
  • Scheduled briefs (Business Agent)
  • Priority compute scheduling
  • Dedicated support
Learn about Max →
Team plans

Scale to your whole team

Tokens and storage multiply per creator. Two creators share one compute pod — you're not paying for idle resources. Viewer accounts don't consume tokens and don't count toward your pod or creator limits.

Contact us for team setup →
Enterprise

For complex deployments

SSO / SAML, on-premise, multi-workspace federation, custom data retention, audit log exports, dedicated SLA, custom integrations — and the Business Agent layer that watches your whole operation 24/7.

Learn more →

Add exactly what you need.

$5 / month per viewer

Viewer account

Read-only access to reports, dashboards, and shared answers. Viewers don't consume tokens and don't count toward your pod or creator limits.

$3 / month per 10 GB

Extra storage

Add storage in 10 GB increments. Useful if you're ingesting large CSVs, IoT time-series, or camera footage archives.

+25% on your plan

Always-on compute

Keeps your instance running 24/7 with no sleep cycle — required for IoT data ingestion, real-time Watch triggers, and camera monitoring that can't tolerate cold-start delays.

Required for IoT + real-time Watch

99.5% monthly uptime SLA. Miss it? Service credit, no questions asked.

Onboarding

We'll set up your first question with you.

Not sure how to connect your first database, or which starter pack fits your team? Book 30 minutes with us. We'll get your first real answer live on your real data — on the call.

Free
while we're in early access
Book a call

Before you start.

How is this different from ChatGPT with a CSV?

ChatGPT gives you an answer and forgets your data when the session ends. Tablize keeps the answer, reruns it next week, and turns it into a dashboard or a watch if you want. It also connects to your real Postgres, MySQL, and APIs — not just uploaded files.

Do I need to know SQL?

No. Tablize writes the SQL for you. If you do know SQL, you can read what it wrote and edit it directly — nothing is hidden behind a black box.

Where does my data live?

On Free, nowhere — your data is analyzed in an ephemeral pod and discarded when the session ends. On Plus and up, it lives in a dedicated Postgres inside your workspace. Self-hosted and on-premise are both supported from Pro upward.

What counts as "enough tokens to finish one real question" on Free?

It's enough for the agent to read your data, write the SQL, run the analysis, draw the chart, explain the result, and handle a few follow-up questions. A real conversation, not a single-shot demo. Exact token budget shows live in the UI as you work.

What's the difference between sleep/wake and always-on?

By default, your Tablize instance sleeps when idle to save compute. It wakes up when you open it or a scheduled job runs. Always-on (+25%) keeps the instance fully running 24/7 — required for IoT devices streaming data continuously, real-time Watch triggers, and cameras that can't tolerate cold-start delays.

Do I need always-on to use Tablize?

No. If you're analyzing spreadsheets, running scheduled reports, or doing periodic analysis, sleep/wake is fine. Always-on is only required if you need real-time event detection — like a sensor firing at 2 AM or a refund spike happening outside business hours.

How does team compute work?

Two creators share one compute pod. A 5-person team runs on 3 pods. This keeps team pricing fair — you're not paying for compute you're not using. Token and storage allocations are fully per-creator, so each person has their own capacity.

Is there an AI model I'm stuck with?

You can bring your own Anthropic or OpenAI key, or use the Tablize-managed pool that ships with every plan. The agent framework is model-agnostic.

Can Tablize work with IoT devices and cameras?

Yes — and this is a category where most BI tools simply can't compete. Tablize connects to MQTT brokers for sensor data and supports camera and video feed ingestion. Enable always-on compute so your agent is awake when the data arrives.

Get started

Your question is waiting.
Your data is ready.

Drop in a CSV. Ask the first question you'd ask a data analyst. See what comes back — and what you'd want to keep.