Prolific Pricing Explained: What 1,000 Responses Cost (2026)
November 9, 2026
Prolific does not publish a price per response. It bills participant time plus a platform fee, so the cost of a study depends on how long each judgment takes. This post works out the cost per 1,000 responses for the tasks AI teams run most: image, text, speech and video comparisons.
We build Datapoint, which competes with Prolific for this work. The Prolific figures below are from prolific.com/pricing, checked November 2026.
The formula
Prolific charges two things:
- Participant reward. You set an hourly rate. The minimum is $8 an hour and Prolific recommends $12.
- Platform fee. 42.8% on top of the reward for companies, 33.3% for academic and non-profit researchers.
So one participant hour costs a company:
| Hourly reward | Platform fee | Cost per participant hour |
|---|---|---|
| $8 (minimum) | 42.8% | $11.42 |
| $12 (recommended) | 42.8% | $17.14 |
The cost of a study is then:
cost = responses x seconds per response x hourly cost / 3600
Cost per 1,000 responses
"Seconds per response" is the whole task: loading the media, playing it, reading the question and voting. Video and speech have to be played before anyone can judge them, so they take longer than images.
| Task | Time per response | At $8/h | At $12/h |
|---|---|---|---|
| Image pair, quick preference | 15 s | $48 | $71 |
| Text or LLM response pair | 30 s | $95 | $143 |
| Speech pair, two 10-second clips | 40 s | $127 | $190 |
| Video pair, two 10-second clips | 45 s | $143 | $214 |
| Video pair with audio, two 20-second clips | 60 s | $190 | $286 |
A typical video evaluation, 500 prompt pairs at 10 votes each, is 5,000 responses: about $1,070 at the recommended rate.
What the formula leaves out
- Instructions and consent. Each participant reads these once, and you pay for the time. Short sessions make it a larger share of the bill.
- The study itself. Prolific recruits the participants. You build and host the task, in your own tool or Prolific's task builder, and aggregate the exported answers.
- Attention checks. Yours to design, and they add time to every session.
- Managed services. Quoted separately.
How this compares with per-response pricing
Per-response platforms such as Datapoint charge one flat rate per response for any modality. The rater's time is not billed, so a 20-second video costs the same to judge as a still image.
On very short tasks the two models come out close. For anything that has to be played, per-response pricing costs about a quarter as much: Datapoint's rate for a video or speech comparison is roughly a quarter of the Prolific figures above.
When hourly pricing is the right model
Paying for time makes sense when time is the product: a 20-minute questionnaire, a behavioural experiment, a longitudinal study that follows the same people for weeks. Prolific was built for that work and is good at it.
For short, repeated judgments of model outputs, time is overhead. See the full Datapoint vs Prolific comparison, or email sales@trydatapoint.com for a quote on your own study.