Choosing a free survey builder looks like a features decision and almost never turns out to be one. Every tool in this category builds a multiple choice question, applies branching logic, and shows a results chart. Where they differ, sharply, is in what they count and what happens when the count runs out.
That difference shows up at the worst possible moment. The survey is live, the link has gone out, and the responses stop being recorded, or they keep being recorded and the account can no longer see them. Or everything works until the day someone needs the raw data in a spreadsheet and finds that export is the thing the paid tier sells.
So the useful comparison is not a feature list. It is four numbers and one rule, per tool, read before the link goes out rather than after.
The four things a free tier meters
Responses. The most common cap, and the one that causes real damage, because it is reached while a survey is running rather than while it is being built. Some tools stop accepting responses. Others keep accepting and stop showing them.
Questions per survey. Less obvious, because it applies at build time. A ten question ceiling does not break anything. It quietly reshapes what the survey can ask, and the constraint is often mistaken for a design decision made freely.
Number of forms or surveys. Matters for teams running several intakes at once, and usually solved by deleting old ones, which trades away the historical data.
Export. The rule rather than the number. Whether the raw responses can leave the tool as a CSV or spreadsheet determines whether a free tier is a place to run a survey or a place to be stuck in one.
Rank those four by which would hurt most in your case before comparing anything, because no free tier is generous on all four and each one gives away a different corner.
One more distinction is worth holding onto while reading any of these numbers. Some caps are per survey and some are per account, and the two behave nothing alike over time. A per survey cap can be reset by starting a new survey, at the cost of splitting the data. A per account cap accumulates across everything built in that account and can only be cleared by deleting past work. Tools that publish a monthly figure and a total figure, as several do, are applying both at once.
What each free tier actually allows
Figures below are taken from each product's own documentation or pricing page in September 2026.
| Tool | Responses on the free tier | Questions per survey | Surveys or forms | Export on free |
|---|---|---|---|---|
| Google Forms | No published cap | No published cap | No published cap | Yes, to Sheets and CSV |
| Microsoft Forms | 200 on a free personal account | Up to 200 | Up to 400 per account | Yes, to Excel |
| Tally | Unlimited, within fair use | No published cap | Unlimited | Yes |
| Jotform Starter | 100 submissions per month, 500 stored in total | Up to 100 fields | 5 active forms | Yes |
| Zoho Survey Free | 100 per survey | Up to 10 | Unlimited | Yes, PDF, XLS and CSV |
| SurveyMonkey Basic | A limited number viewable per survey | Up to 10 | Not the binding limit | No |
The spread on responses runs from 100 to unlimited, and the spread on questions runs from 10 to no stated ceiling. Two tools in that table are from the same category of vendor and differ by a factor of large enough that they are not really competing for the same job.
The response cap is the one that bites mid-project
Microsoft Forms is the clearest illustration, because the same product carries wildly different limits depending on the account signed in.
The published limits page states up to 200 responses per form on a free personal account, up to 1,000 on a paid personal account, and up to 5,000,000 on a Microsoft 365 work or school account. It also lists up to 400 forms per account and up to 200 questions per form, with a 4,000 character limit on a question and on a single response.
The practical consequence is that advice about Microsoft Forms is unreliable unless it says which account type it applies to. Someone running a survey from a work tenant has effectively no response ceiling. Someone using the same interface with a personal address has 200, which a single mailing list can exhaust in an afternoon.
Jotform's Starter plan meters differently again, with 100 submissions per month and 500 total submission storage, alongside 5 active forms, 10,000 monthly form views and 100 MB of upload space. The monthly counter resets, the total storage counter does not, so a form that runs for six months hits the second ceiling even while staying comfortably under the first.
Zoho Survey's free plan allows unlimited surveys with 100 responses each and a single user, which suits several small independent surveys better than one large one.
Tally's free plan takes the opposite approach and places no cap on forms or submissions, subject to fair use, with the product's own branding shown on forms and removal reserved for the paid tier. Google Forms similarly publishes no response ceiling and exports directly into Sheets.
Reading the unlimited tiers honestly
Unlimited is a real advantage and it is not the same as unconstrained. Where a tool gives away volume, it charges for something else, usually branding removal, custom domains, team seats or data handling controls. That is a legitimate trade and it is worth knowing which one you are accepting. A survey sent to customers with another company's name on it is a different decision from a survey sent internally.
The question cap shapes the survey before you notice
A ten question ceiling, as applied by SurveyMonkey's Basic plan and Zoho Survey's free plan, is not a small constraint. It is roughly enough for a satisfaction survey with a rating, a reason, and a few segmentation questions, and it is not enough for anything diagnostic.
What makes it worth flagging is that the limit is invisible in the results. Nobody looking at the finished survey can tell that three questions were cut to fit, and the analysis proceeds as though the survey asked everything it needed to. A response cap announces itself. A question cap silently narrows the conclusions.
If the survey genuinely needs more than ten questions, that decides the shortlist before any other consideration, and the tools with no stated question ceiling are the only real candidates.
Export is where the line actually falls
This is the single rule most worth checking before sending a link, because it is the one that cannot be worked around afterwards.
SurveyMonkey's individual plans page lists the Basic plan as free with a maximum of 10 questions per survey, a limited number of responses viewable per survey even when more are collected, and data exports not included. Responses continue to be gathered. Seeing all of them, or taking any of them out as a file, is what the paid tiers provide.
The other tools in the table include export on their free tiers. Zoho Survey's free plan lists PDF export and reports in XLS and CSV. Google Forms writes responses into a Sheet. Microsoft Forms exports to Excel. Jotform and Tally both allow the data out.
Export availability is also the best available proxy for how easy it will be to leave later. A tool that lets the raw data out on a free tier is one where the responses remain yours in any practical sense. A tool that does not is one where the survey has to be rerun elsewhere if the relationship ends, and rerunning a survey does not recover the people who already answered.
Costs that do not appear in any plan table
Three things regularly decide the choice and appear in none of the numbers above.
Whether respondents have to sign in. This is the quietest killer of response rates, because the people who bounce never appear in the data. Google's documentation for the file upload question type states that responders need to sign in to a Google Account to answer it, that uploaded files land in a folder on the form owner's Drive, and that the question cannot be used if the form is stored in a shared drive. Other settings in the same product, such as verified email collection or limiting to one response per person, have the same effect. A survey sent to the general public should be answerable by someone with no account anywhere.
Whose account the survey belongs to. Free tiers are single user by design. Zoho Survey's free plan lists one user. Jotform's Starter plan is one account. In practice this means the survey belongs to whoever set it up, and when that person leaves or is on holiday, the responses go with them. Teams work around it by sharing one login, which is worse than it sounds for anything containing personal data.
What the branding says. Where a tool gives away unlimited volume, it usually reserves branding removal for the paid tier, as Tally does. Internally that is irrelevant. For a survey sent to customers, partners or job applicants, another company's name on the page changes how the request reads, and that is a judgement rather than a cost.
The pattern across all three is the same. Free tiers are priced for one person running an occasional survey, and the friction appears at exactly the point where a survey becomes a recurring part of how a team works.
What the free tier is not measuring at all
Every number above concerns collection. None of them concerns what happens next, and for a large share of the forms people build, that is where the real work sits.
A satisfaction survey produces an average. An application form, a support request or an enquiry produces a queue of individual cases, each of which needs someone to own it, a state showing whether it has been dealt with, and a reply that goes back to a specific person. Survey tools are built for the first shape and are borrowed for the second, which is why so many teams end up exporting to a spreadsheet, adding columns for owner and status by hand, and then losing track of which export is current.
If the responses need replies rather than charts, the metering question changes. What matters is whether responses are unlimited regardless of volume, and whether the tool carries owners, stages and internal notes without a spreadsheet in the middle. That is a different product category, and it is worth looking at what response handling involves and the situations it fits before committing a process to a survey tool.
What to change first
Estimate the number of responses realistically, count the questions the survey actually needs, and check the export rule, in that order, before the link goes out. If any of the three is close to a free tier ceiling, choosing differently now costs nothing and choosing differently mid-survey costs the responses already collected. If the answers will need replying to rather than counting, look at a tool that keeps forms and responses unlimited and prices only on how many people use it, such as Halict.
Q1. Which free survey builder has no response limit?
Google Forms and Tally both publish no cap on the number of responses on their free tiers, with Tally applying a fair use guideline and showing its own branding on forms. Microsoft Forms has no practical ceiling when used with a Microsoft 365 work or school account, where the published limit is 5,000,000 responses per form, but only 200 on a free personal account.
Q2. Why does Microsoft Forms stop at 200 responses for some people?
Because the limit depends on the account, not the product. Microsoft's published limits list up to 200 responses per form on a free personal account, up to 1,000 on a paid personal account, and up to 5,000,000 on a work or school account. The interface is the same in each case, which is why advice about the limit is often contradictory.
Q3. Can survey responses be exported for free?
It depends on the tool. Google Forms, Microsoft Forms, Zoho Survey, Jotform and Tally all allow data out on their free tiers. SurveyMonkey's Basic plan lists data exports as not included and shows only a limited number of responses per survey. Checking this before sending the link matters, because responses already collected cannot be moved later.
Q4. Is a 10 question limit enough for a customer survey?
For a short satisfaction check with one rating, one open comment and a few segmentation questions, usually yes. For anything diagnostic, where the goal is to understand why a score is what it is, ten questions forces cuts that are invisible in the final results. The limit applies on SurveyMonkey's Basic plan and Zoho Survey's free plan.
Q5. When is a survey tool the wrong choice entirely?
When each response needs a reply rather than a count. Applications, enquiries, support requests and registrations produce a queue of individual cases that need an owner and a status, and survey tools are built to aggregate rather than to track. The usual symptom is a spreadsheet export with columns added by hand for who is handling what.