| What it does | Human evaluation for text-to-speech with blinded listening studies
| AI release readiness layer for software built by humans and AI agents
|
|---|
| Stage | Not given
| Startup
|
|---|
| Founded | Not given
| Not given
|
|---|
| Based in | US
| IN
|
|---|
| Pricing model | Paid
| Not given
|
|---|
| Pricing | $18.00 per hour of audio heard by listeners; $0.30 per judgement on a pair of two 30-second clips; $300.00 for 200 pairs × 5 ratings with 30-second clips (1,000 judgements)
| Not given
|
|---|
| Key features | - Blinded paired-preference studies with no system names visible
- API for creating studies and retrieving results
- Exact price quote before study starts
- Per-judgement data export as CSV
- Paid US-based listener panel
- Attention signals recorded with every judgement
- Live results streaming as listeners answer
- Detection template for AI vs human narration
- Published methodology and platform transparency
| - Chat-based test creation and execution in TestNeo Studio
- URL scanning to generate plain-English test suggestions
- Browser recorder for capturing user journeys
- Visual regression testing with smart diffs
- API test generation and risk discovery from Swagger
- Release intelligence with PASS/WARN/BLOCK verdicts
- Agent verification comparing claims against observed behavior
- Integration with Jira, Figma, Swagger, and incident tracking
- Local agent option for apps behind VPN
- k6 load testing insights and export
|
|---|
| What makes it different | - Publishes its own listening studies on the same platform with the same methodology as customer studies
- Forced-choice preference testing answers which system people prefer, not rating scales
- Fast turnaround with live result streaming and no minimum study size
- Transparent pricing with no setup fees or minimums, quoted before starting
- Customer audio stored securely and used only for that study
| - Plain English test creation without coding or Playwright scripts required
- Predictable AI costs with rule-based replay instead of per-step model calls
- One integrated platform for all testing types: web, API, visual, load, and agent verification
- Release evidence with comprehensive test results, verdicts, and reasons in one place
- Bring your own AI model at same seat price with no markup on tokens
|
| Community votes | 0 | 0 |
| DR (Domain Rating by Ahrefs) | 10 ▲ +2 (up) since last check | 11 ● no change since last check |
| Trust Flow (Majestic) | 9 | 5 |
| Citation Flow (Majestic) | 24 | 17 |
| Referring domains (Majestic) | 69 | 25 |