Which Survey Platforms Support Voice Responses in 2026

Two different questions hide inside one

“Can my survey accept a spoken answer?” has two answers, and people usually get the wrong one because they only ask half the question.

1. Does the platform record voice itself? A built-in question type where the respondent taps a button and speaks. No third party involved.
2. Can the platform host somebody else’s recorder? That means running your own JavaScript on the page the respondent sees. Some platforms allow it; several deliberately do not.

A platform can say yes to one and no to the other, and the two paths have completely different consequences for your study. The table below separates them.

The compatibility table

Every row was checked against the platform’s own documentation in September 2026. Where a platform changes, this page is wrong until we re-check it — so the sources are linked rather than summarised away.

Typeform: use the native feature

Typeform ships a Video and Audio question type. The respondent can answer by video, by audio, or by text if you leave text enabled; Typeform generates a transcript automatically and lets you download the recording. Two limits worth knowing before you plan a study around it: each answer is capped at 2 minutes, a limit their documentation says cannot be adjusted, and the feature requires a Growth or Talent plan. Automated captions are English-only at the time of writing.

If you are on Typeform and on one of those plans, this is the right answer and you can stop reading. You do not need us, and it would be dishonest to pretend otherwise. Their own guidance is worth following too: offer voice as an option rather than forcing it.

SurveyMonkey: the documentation people find is about something else

Search SurveyMonkey’s help centre for audio and you land on Adding Video or Audio, which covers embedding a YouTube, Vimeo or SoundCloud player into a survey. That is playback — media the respondent listens to. It is not a way for the respondent to record an answer, and the two get confused constantly because they share a word.

SurveyMonkey also strips custom JavaScript from the respondent’s view, so the widget route is closed as well. If you are on SurveyMonkey and you need spoken answers, you are choosing between platforms, not between features.

Alchemer, QuestionPro, FormAssembly, Sawtooth and Toluna Start: the widget route

None of these record voice natively, and all of them let you run your own JavaScript on the respondent page. That is the opening a third-party recorder needs: the microphone lives inside the question, the study stays where it is, and the transcription is written back into the field so it arrives in your normal export next to every other answer. Qualtrics also accepts custom JavaScript; Voice Capture's support for it is coming soon.

Toluna Start is the newest row. It already works on open ends in text, with QProbe asking an AI follow-up when an answer is thin, but it has no voice answer type. Its ADD SCRIPT option runs code on each question, which is where the recorder goes; we tested it end to end in October 2026 and the transcript appeared in Toluna’s Analyze view. The walkthrough is how to add voice responses to Toluna Start surveys.

What to do when your platform can’t

If the answer is no on both counts, there are only three architectures, and it is worth being clear-eyed about what each one costs.

The first is the one most comparison posts recommend, and it is the one most teams cannot actually take. A survey platform is an annual licence, and the studies programmed inside it are years of somebody’s work that does not travel. Switching to get a better open end means writing both off — which is why the better open end usually never gets tested at all.

Where we sit, stated plainly

Voice Capture is the widget in row three: a block of code you paste into a survey you already field, on any platform that lets you add your own HTML and JavaScript. Respondents speak, the answer is transcribed automatically in 36 languages and written back into the field. Typing stays available throughout.

Which means the honest version of this page is that we are the wrong answer for two of the platforms above. On Typeform, use their native question. On SurveyMonkey, we genuinely cannot help you without a platform change. We would rather say that here than after you have signed up.

If you are on Alchemer, QuestionPro, FormAssembly, Sawtooth, Toluna Start or anything else that runs your own JavaScript, the setup guides are at /integrations, and the measurements behind our claims — including what they cannot tell you — are at /methodology.

Sources

• Typeform Help Center, Collect video and audio answers from respondents — the 2-minute cap, the Growth/Talent requirement and the automatic transcript.
• SurveyMonkey Help Center, Adding Video or Audio — embedding YouTube, Vimeo and SoundCloud players into a survey.
• Sawtooth Software, Discover Help: Custom JavaScript — custom code and the addOnQuestionLoadHandler hook. Which outside domains that code may reach is set by the Content Security Policy header of Discover’s survey pages, which Sawtooth support maintains; their support team also confirmed that Discover has no voice question of its own. Checked 5 October 2026.
• Toluna, Toluna Start platform page — QProbe (AI follow-ups on open ends) and SmartCloud (open-end coding); no voice or audio answer type is listed. ADD SCRIPT and the transcript in Analyze come from our own end-to-end test. Checked 24 October 2026.
• Checked 22 September 2026. Platforms ship things; if you are reading this much later, verify before you plan around it.