Remesh, Synthetic Users and Simile are often compared, but only two of them are synthetic. Remesh collects answers from real people in live or asynchronous group conversations and uses AI to summarise them. Synthetic Users simulates one-to-one interviews with AI participants. Simile simulates how customers, employees or populations will respond to a change. Pick Remesh when the answer has to come from real people, Synthetic Users for fast synthetic interviews in product and UX work, and Simile for enterprise behaviour simulation. Facts below come from each vendor's website, checked on October 4, 2026.
Minds is our product and a fourth option: the end-to-end platform for commercial synthetic research, where one team builds Audiences and runs qualitative and quantitative Studies on them. We include it where it fits and state its limits.
| Dimension | Synthetic Users | Remesh | Verdict |
|---|
| Who answers | AI participants built from an audience description and your data | Real customers or employees | Remesh when the answer must come from real people |
| Format | One-to-one synthetic interviews plus an insights report | Live, asynchronous or video group conversations with many participants at once | Synthetic Users for individual depth; Remesh for group consensus |
| Recruitment | None | Your own audience or Remesh's vetted participant recruiting | Synthetic Users when recruiting is the bottleneck |
| Pricing | Annual plans from $12,500 a year; $2 to $60 per interview | Not published; request a demo | Synthetic Users is easier to budget |
| Accuracy evidence | 85% to 92% parity with real interviews on themes, in vendor studies | Answers come from real people, subject to sample and moderation bias | Use synthetic to narrow, Remesh to confirm |
| Best for | Product discovery, concept and UX interviews before recruiting | Consumer insights at scale, employee listening, healthcare research | Different stages of the same decision |
Synthetic Users describes itself as user research without recruitment: you define the participants, ask your questions, and AI participants answer in interview form. Studies can be grounded in your own transcripts and support tickets, and each run ends in an insights report. Plans are annual, start at $12,500 a year on a pool of research tokens with seats, and the vendor quotes $2 to $60 per interview against $100 or more through traditional agencies (pricing). Its homepage reports 85% to 92% synthetic-organic parity, measured on thematic overlap, depth and qualitative alignment (source). Theme overlap shows that synthetic interviews raise similar topics; it does not show that answer shares match a real population. Research responses consume the prepaid balance or the applicable paid-plan allowance.
Remesh runs conversations with real people. A researcher builds a discussion guide, invites an audience (customers, panel respondents or employees, or participants recruited through Remesh's own vetted recruiting), and hundreds of people answer at once in a live, asynchronous or video session. AI groups and ranks the open-ended answers while the session runs, and the Remy agent helps with research design and analysis (source). Remesh sells through a demo and does not publish prices. Because every answer comes from a person, Remesh carries the usual human-research costs and biases: incentives, scheduling, and the way a group conversation can pull answers toward the majority.
If you searched for Simile vs Remesh, the key point is that they solve different problems. Simile is a synthetic platform: it trains a model of human behaviour on large proprietary studies and simulates how customers, employees or populations respond to a change, with a predicted confidence attached to each result (source). It raised $200 million at a $2 billion valuation and names CVS Health, Wealthfront, Deloitte and Gallup as customers (source). Remesh does not simulate anyone; it gathers real answers.
For synthetic user testing, then, Remesh is the confirmation step rather than an alternative to Simile. A practical sequence is to simulate first (in Simile, Synthetic Users or Minds), narrow the options, then run the shortlisted concepts past real people in Remesh. Simile is enterprise only and does not publish prices, so smaller product teams usually compare it with self-serve tools instead. For a deeper look, see Simile alternatives and Minds vs Simile.
Both are synthetic, but they target different buyers. Simile aims at consequential enterprise decisions and reports a confidence per simulation; the research it builds on, Generative Agent Simulations of 1,000 People, reproduced General Social Survey answers at 83% to 86% of participants' own two-week retest consistency (source). That is an academic benchmark, not a test of the commercial product. Synthetic Users aims at product and UX teams and returns interview transcripts and a report. Choose Simile when the question is how a population will behave; choose Synthetic Users when you want qualitative interviews about a product idea. See also Synthetic Users alternatives.
Some buyers add Suzy to the shortlist. Suzy now sells the Suzy Decision Engine, a consumer-insights platform that brings market intelligence, your data and documents, and stakeholder deliverables together (source). Compare it in Minds vs Suzy.
Choose Synthetic Users when a product or UX team wants interview-style feedback before recruiting anyone: early problem discovery, reactions to a concept or prototype, or preparing an interview guide. It suits teams that think in transcripts and themes and can commit to an annual plan.
Choose Remesh when the decision needs real voices: consumer insights you will quote, employee listening, healthcare research, or a check on concepts a synthetic study has already narrowed. It fits organisations that can recruit or invite participants and work in scheduled or asynchronous sessions.
Minds is the end-to-end platform for commercial synthetic research. Instead of one interview format, you build a reusable Audience of Minds (AI personas) grounded through Minds PRISM in public sources and the files and briefs you add, then run Studies on it: one-to-one interviews, moderated discussions between several Minds, or questionnaires with open-ended, single choice, multiselect and scale questions. Eleven deterministic methods run inside a Study, including MaxDiff, NPS, TURF, Van Westendorp and conjoint, so qualitative and quantitative work stays in one workflow. You can test copy, websites, images, video, decks and Figma inputs where enabled.
Each Audience can be validated against real published surveys or your own survey files, with a score out of 100 per survey, so you can see where the simulation tracks real data and where it does not. Pricing is published: Pay as you go at $0.12 / €0.12 per response and Pro at $199 / €199 per seat a month.
Minds' limits: answers are simulated and directional, not a representative sample. Minds has no recruited panel, so when you need real people, pair it with a human platform such as Remesh. Read the methodology for how Audiences are built and checked.
Synthetic tools remove recruitment and scheduling, so a team can test many variations in a day. Remesh sessions are quick once participants are in the room, but the full cycle includes guide design, recruitment or invitations, incentives and analysis, and a second round means a new session. In cost terms, Synthetic Users and Minds publish prices, while Remesh and Simile quote after a demo. Synthetic answers are cheaper per answer, but they are only worth it if you check them: rerun a past study where you know the real answer before you rely on any synthetic tool for that kind of decision.
Remesh processes personal data from real participants, so consent, privacy notices and a data processing agreement are part of every study. Synthetic tools do not survey real people during simulation, but the concepts, customer files and notes you upload can still be confidential or personal data. Minds offers a data processing agreement under GDPR Article 28 and publishes its subprocessors; ask every vendor for the equivalent and review it with your privacy team.
Synthetic answers, from any vendor, should not be the only evidence for:
- Final go or no-go decisions on a large launch or investment.
- Statistically representative estimates such as market size or share.
- Physical or sensory products, where taste, smell or handling matter.
- Regulated claims, such as health or financial claims.
- Behaviour measurement, such as clicks, conversion or purchase.
In those cases, run the study with real people, in Remesh or another human research platform, after synthetic work has narrowed the options.
For a wider view of synthetic vendors, see synthetic respondent platforms compared, Minds vs Synthetic Users and Remesh vs Koji.
Synthetic Users and Remesh are not substitutes. Remesh is the right tool when real people must answer, at scale and in a group. Synthetic Users is the faster, cheaper first read for product and UX interviews, and Simile is the enterprise option for behaviour simulation at population level. If one team needs synthetic qualitative and quantitative research in a single workflow, with each Audience checked against real surveys, try Minds, then confirm the decisions that matter with real people.
