Sigma Computing Pricing: Sigma Computing Cost, License Tiers and Competitors
Sigma's website shows no price at all. Its own AWS Marketplace listing does: a $150,000 platform fee and $2,000 per Creator. Here is what that means for your budget, and what 153 real buyers ended up paying.
Sources: Sigma's AWS listing, Sigma docs, Vendr · nothing estimated
Ask the demo shop a question, or type your own:
Reading the schema, writing SQL… Writing SQL… Running (read‑only)… SQL Agentsql wrote
▋
A real MySQL shop: customers, products, orders and signups over the last year. Agentsql writes the SQL, runs it read-only, and answers.
Direct answer
Sigma Computing does not publish pricing on its website: sigmacomputing.com/pricing now redirects to a sales contact form. The only public Sigma price list is on Sigma's own AWS Marketplace listing, which prices a 12-month contract at $150,000 for the Enterprise Platform with 24/7 support, $2,000 per Creator, $2,000 per Explorer, and $0 per Viewer. Real contracts come in well under that: Vendr's data on 153 Sigma purchases shows a median of $65,080 a year, ranging from $17,060 to $146,240. On top of the Sigma bill you pay your warehouse vendor, because every Sigma query runs live on your Snowflake or BigQuery compute. We read all of these sources on 17 September 2026.
›_ sigma computing pricing, the one public list
The price Sigma publishes, just not on its own website.
Sigma's AWS Marketplace listing, seller "Sigma Computing", 12-month contract. Quoted exactly.
Every article on page one of Google answers "how much does Sigma cost" with a range and a shrug, because Sigma's pricing page is gone. Type the URL and you get an HTTP 308 redirect to a contact form. We counted the dollar signs in that form's raw markup and every one of them is a JavaScript framework reference, not a price. But Sigma sells through AWS Marketplace too, and a Marketplace listing has to show a contract table. Sigma's does.
| Dimension | Sigma's description | Cost per 12 months |
|---|---|---|
| 14-day Free Trial | Free Trial with 5 Maximum Users | $0.00 |
| Enterprise Platform | Enterprise Platform + 24/7 Enterprise Support | $150,000.00 |
| Add On, Creator | Admin access, create joins, worksheets, dashboards, write back | $2,000.00 |
| Add On, Explorer | Drill down dashboard to worksheet and data sets (curated by Creators) | $2,000.00 |
| Add On, Viewer | Interactive Dashboards with filtering capabilities | $0.00 |
What it tells you
Sigma is priced as a platform first
The seat prices are small next to the platform line. At list, the $150,000 fee is the same as 75 Creator add-ons. Headcount matters much less to a Sigma bill than whether you are on the platform at all.
What it does not tell you
How many users the fee includes
The listing does not say whether the platform fee bundles any Creators. It also still uses the older Creator, Explorer and Viewer names, while Sigma's documentation has moved to four tiers. Ask sales which model your quote is written on.
One trap in the search results
There are two "Sigma" listings on AWS
Search AWS Marketplace for Sigma and you will also find a sensor vendor selling $50,000, $100,000 and $250,000 bundles. That is a different company. If a guide quotes those figures as BI pricing, it grabbed the wrong listing.
›_ sigma computing licenses, current model
View, Act, Analyze, Build: the four tiers Sigma bills on now.
From Sigma's documentation page "Account type and license overview". The "cannot" column is ours, read off Sigma's permission matrix.
Most pricing guides still describe Sigma as Essential, Business and Enterprise plans, or as Creator, Explorer and Viewer seats. Sigma's own docs describe four license tiers, and they bill each account type at the highest tier of any permission it has switched on. That last detail matters: give one Analyze user a single Build permission, such as writing SQL, and that user now costs a Build license.
| Tier | Built for | What it can do | What it cannot |
|---|---|---|---|
| View | Report consumers who need prepared data and insights | View workbooks and reports, use controls and drill paths, download PDF and PNG, comment, and use Sigma Assistant to ask natural language queries | Cannot filter underlying data, schedule exports or do ad hoc analysis |
| Act | Contributors who input and update data | Everything in View, plus forms and input tables, stored procedure and API actions, full workbook downloads, scheduled exports and alerts | Cannot perform ad hoc analysis |
| Analyze | Decision makers who dig deeper without building | Everything in Act, plus ad hoc analysis and the AI formula assistant | Cannot create workbooks, model data, write SQL or see Custom SQL logic |
| Build | Data architects, BI analysts and report builders | Everything in Analyze, plus create and share workbooks, model and materialize data, manage users and connections, write SQL and Python, view Custom SQL logic | Nothing held back. Admins also need this tier |
Two things in that table are worth carrying into a negotiation. First, Sigma put natural-language questions (Sigma Assistant) into the lowest tier, which is generous and means a View user can ask a question in English. Second, only Build users can see Custom SQL logic. Everyone below Build gets the answer without the query that produced it, so when a number looks wrong, the check has to go back to a Build user. The docs also state plainly that they are "general product information only" and that license pricing comes from your Account Executive.
›_ sigma computing cost, worked example
What a 183-person rollout costs at Sigma's list price.
Arithmetic on the AWS listing only. Your warehouse bill is separate and comes next.
| Line | Qty | Per year | Total |
|---|---|---|---|
| Enterprise Platform, 12 months | 1 | $150,000 | $150,000 |
| Creator add-on | 8 | $2,000 | $16,000 |
| Explorer add-on | 25 | $2,000 | $50,000 |
| Viewer add-on | 150 | $0 | $0 |
| List total, before warehouse compute | 183 people | $216,000 |
Take a mid-market analytics setup: 8 people building workbooks, 25 managers who drill into data, and 150 people who only read dashboards. At Sigma's AWS list that is $216,000 a year, and the 150 readers add nothing to it.
Now set that against what buyers actually sign. Vendr's median Sigma contract is $65,080, its top observed deal is $146,240, and it reports buyers saving 25 percent on average. So the AWS listing reads best as a ceiling. Hardly anyone pays it, and it is useful precisely because it is the only published anchor you can put in front of a sales rep.
Read the Viewer line again, too. $0 per reader is a real strength, and for a company whose main problem is getting dashboards in front of hundreds of people, it is a good reason to shortlist Sigma.
›_ sigma computing snowflake cost
The second bill: every Sigma query runs on your warehouse.
Sigma stores nothing. Your Snowflake, BigQuery, Databricks or Redshift account does the work and bills you for it.
Sigma's design is live queries against the warehouse, with no extracts. That is why its numbers are always fresh, and it is also why a Sigma rollout shows up on a second invoice. Each dashboard open, filter change and drill-down sends SQL to your warehouse. On Snowflake that consumes credits for as long as the virtual warehouse runs; on BigQuery on-demand it bills the bytes each query scans.
This is the line that free Viewer seats make easy to underestimate. 150 readers cost Sigma nothing, but 150 readers each opening a morning dashboard with six charts is roughly 900 warehouse queries before 10am. How much that costs depends on caching, warehouse size and auto-suspend settings, so price it by running your real dashboards during the 14-day trial and reading the query history. Our Snowflake pricing and BigQuery pricing pages have the current per-credit and per-TiB rates to plug in.
›_ sigma computing competitors, by price
Sigma Computing competitors, and what each one actually publishes.
List prices as published by each vendor. Where a vendor publishes nothing, we say so.
| Tool | Published price | Pricing model | Best for |
|---|---|---|---|
| Sigma Computing | No price on its website. AWS listing: $150,000 platform plus $2,000 per Creator or Explorer a year, Viewers $0 | Platform fee plus role add-ons, queries run live on your warehouse | Snowflake or BigQuery shops that want spreadsheet-style exploration and write back |
| Tableau Cloud | Creator $75, Explorer $42, Viewer $15 per user per month, billed yearly | Per role seat | Teams that want the deepest visualization tooling |
| Power BI | Pro $14, Premium Per User $24 per user per month; Fabric capacity from F2 | Per seat, or capacity | Microsoft 365 organizations |
| ThoughtSpot | Essentials and Pro tiers published per user; Pro $50 per user per month | Per user, enterprise by quote | Search-style analytics at enterprise scale |
| Looker | No dollar price for any edition | Platform edition plus user types, by quote | Google Cloud shops with a LookML team |
| Agentsql | Starter $39, Team $99, Scale $299 per month, billed yearly | Flat team plan, no platform fee | Teams that need plain-English answers from Postgres, MySQL, Snowflake or BigQuery |
The detailed rate cards live on their own pages: Tableau pricing, Power BI pricing, ThoughtSpot pricing and Looker pricing. If you are weighing Sigma directly against Tableau on features rather than cost, Sigma vs Tableau covers the live-warehouse versus extract trade-off in full.
›_ honest fit
When Sigma is the right buy, and when a $65,000 contract is the wrong tool.
Choose Sigma if
Your data already lives in Snowflake or BigQuery, you have analysts who will build and govern workbooks, and you want hundreds of business users exploring live data in a spreadsheet grid they already understand. Input tables and write back are real strengths, as are $0 Viewers and natural-language questions in the lowest tier. At that scale, a five-figure contract buys a lot.
Choose something lighter if
What your team actually needs is answers, not a BI platform. If the requests pile up as "how many accounts churned last quarter by plan" and nobody wants to build a workbook for each one, you are paying for a modeling layer to solve a question-asking problem. Vendr's purchase data puts a typical contract near $65,000 a year before warehouse compute.
That second case is what Agentsql does. It connects read-only to your Snowflake, BigQuery, Postgres or MySQL database, turns a question in plain English into SQL, runs it, and returns a chart, a table and a one-line answer. Every user sees the SQL, not just the ones on the top tier, so anyone can check the logic. There is no platform fee and no modeling project to finish before the first answer. The published price is $39, $99 or $299 a month billed yearly, which means the Scale plan for a full year costs about 5.5 percent of Vendr's median Sigma contract.
We are not a spreadsheet interface and we do not do write back or input tables, so if those are the reason you shortlisted Sigma, keep it on the list. If the reason was "business people keep waiting on the data team", try the console at the top of this page against your own schema first. If you are still at the shortlist stage, our self-service BI software guide compares the approaches side by side.
›_ how to price sigma before the call
Four steps to a Sigma quote you can compare.
01
Map people to tiers
Count who needs View, Act, Analyze and Build. Watch for users who need one Build permission, because it bumps them to Build.
02
Anchor on the AWS list
Price your mix at $150,000 plus $2,000 per Creator or Explorer. That is your ceiling, and it is public.
03
Measure warehouse load
Run real dashboards in the 14-day trial and read the query history in Snowflake or BigQuery. Multiply out to a month.
04
Ask for unit rates
Request the per-license price and the platform line separately, plus a renewal cap. A single total is impossible to compare next year.
›_ frequently asked
Sigma Computing pricing questions, answered.
How much does Sigma Computing cost?
Sigma publishes no price on its website; its pricing URL redirects to a contact form. The only public list is Sigma's own AWS Marketplace listing: $150,000 a year for the Enterprise Platform with 24/7 support, plus $2,000 a year per Creator or Explorer, with Viewers at $0. Vendr puts the median contract at $65,080 a year across 153 purchases.
Does Sigma Computing have a free trial?
Yes. Sigma's AWS Marketplace listing includes a 14-day free trial capped at five users, priced at $0.00. That is a short window for a tool whose cost depends heavily on warehouse compute, so use it to measure how many queries your real dashboards fire against Snowflake or BigQuery, not just to click through the interface.
What are Sigma Computing license types?
Sigma's documentation lists four license tiers: View, Act, Analyze and Build. Each account type is billed at the highest tier of any permission it holds. Build is the only tier that can create workbooks, model data, write SQL and view Custom SQL logic. Organizations on the older Lite, Essential and Pro model are pointed to a separate page.
Is Sigma cheaper than Tableau?
For large read-only audiences it can be, because Sigma lists Viewers at $0 while Tableau charges $15 per Viewer a month. For a small team it usually is not: Sigma's list platform fee alone is $150,000 a year, which buys about 166 Tableau Creator seats. Sigma also runs every query on your warehouse, so add that compute to its side.
Does Sigma charge for Snowflake compute?
No, your warehouse vendor does. Sigma does not store your data; it sends queries to your Snowflake, BigQuery, Databricks or Redshift account, and those queries consume credits or slots billed by that vendor. It is a real line on the total cost of Sigma, and it grows with how many people open live dashboards and how often.
What are the main Sigma Computing competitors?
The tools buyers most often weigh against Sigma are Tableau, Power BI, Looker and ThoughtSpot, plus Hex and Omni for warehouse-native analytics. Tableau and Power BI publish seat prices, ThoughtSpot publishes two tiers, and Looker publishes none. For teams that mainly need plain-English answers from a database, an AI data analyst is the lighter alternative.
Can you negotiate Sigma Computing pricing?
Yes, and most buyers do. Vendr reports buyers saving 25 percent on average across 153 Sigma purchases, and its median contract of $65,080 sits far below the $150,000 platform fee on the AWS listing. Multi-year terms, a precise Creator count, and a competing quote are the usual levers. Ask for the per-license rate, not only a total.
›_ pricing and comparisons
›_ connect your database