ChartPixel
Upload a spreadsheet and get charts with a written statistical reading, chat and PowerPoint export
Upload a spreadsheet and get charts with a written statistical reading, chat and PowerPoint export
Fast-read signals for fit, pricing, and trust.
A tidy table turned into presentable charts with an interpretation you can quote is the job here, and the plans are sized by dataset rather than by user: Free is a one-project look, Silver and Gold differ by file size (10,000 versus 100,000 rows), team seats appear only on Gold, and API calls come with Gold, Business or a credit pack. Before paying, note that the pricing page prints the same AU$15.00 figure against both Silver and Gold, and that nothing published says how long an uploaded file is kept.
Sign-up asks for your role and how proficient you are in data analysis, and the workspace is built around projects: one upload becomes one project holding the cleaned dataset and a set of generated chart cards. Each card carries a title, a description and an insight, and the vendor's own worked example shows the insight naming a Cuzick test and flagging which category differences are not statistically different. From there you either take the charts the tool chose or switch to Data Exploration to pick columns and chart types yourself, snapshot the ones you like, and export.
Paying works two ways, and they suit different rhythms. A subscription raises the row, column and prompt ceilings and, on Gold, adds team seats and an API allowance; a credit package is pay-as-you-go, where one credit unlocks one project and the same credits double as API allowance, at one credit per hundred successful calls. Credits do not expire, which makes the pack the better fit for work that arrives a few times a year, while a subscription is what pays for itself on a weekly reporting cycle. On throughput, unsubscribed API users are limited to ten calls a minute and subscribed users to fifty, and a /chart response takes a few seconds rather than being instant — which matters if you intend to call it inside a page load rather than on a schedule.
Quick fit check against how you actually work.
What you can actually do with this tool.
Each chart is generated with a title, a description of what is plotted and an insight paragraph that says which differences are significant, which could be chance, and which test was used. That is the part you would otherwise have to run yourself before putting a claim in front of an audience.
Questions can be asked of the dataset or of a specific chart in plain language. The prompt allowance is where the plans differ most sharply, so check the plan cards before relying on it as a daily workflow.
Charts and their insight text export together to PowerPoint or presentation slides, and finished charts can be embedded or shared by link, so the analysis does not have to be re-typed into the slide that presents it.
Uploads pass through automated data cleaning before any chart is drawn; a guided exploration mode lets you pick the columns yourself; zip codes become maps and free-text comments get sentiment analysis, which stretches the tool beyond numeric columns.
Point the /chart endpoint at a csv, xlsx, feather or parquet file or a plain HTML table (at most 10 MB), name one or two columns and optionally a filter and chart kind, and the JSON response carries the chart as svg, png or html with its generated title and insight. The /tiny endpoint skips the insight for speed; the vendor quotes 4 to 6 seconds for a /chart call.
The vendor says it does not train on your data or send raw data back and forth to external LLMs, and its API documentation says API data input is not stored. For a team that is otherwise pasting commercial figures into a general chatbot, that is the substantive difference — it is a statement about what happens to the numbers, not a feature of the charting.
Pricing tiers and what's included in each.
Common questions about this tool, answered.
On the current pricing page Silver and Gold are annual subscriptions listed at AU$15.00 per month billed annually, each carrying 'Get 5 months free', and the credit package is a one-time AU$15.00 for 5 credits. That one amount stands against both tiers even though their limits differ sharply, so confirm which of them it belongs to before you pay. Business is listed from US$250.
You must sign up, then hold either credits or a subscription. The pricing page assigns API calls to Gold, Business and credit packs only. The API documentation, dated 2019-06-01, separately states that subscribed users can make up to 5,000 API calls per month and can buy credits at a 50% discount once they hit it; because that does not match the per-plan figures, treat the pricing page as the entitlement and confirm with the vendor if the API is your main use.
The home page states that ChartPixel neither trains on user data nor passes raw data to external LLMs, and the API documentation states that API data input is not stored while the output chart can be kept temporarily or permanently as you choose. Column encryption is excluded from the Free plan. What those statements do not cover is a retention period, a deletion process, sub-processors, hosting location or any certification, so ask for those directly if the datasets are sensitive.