The generated presenter stayed professional and consistent, with lip-sync reported as consistent in the finished video.
What was measured
Avatar realism
How believable the presenter looks and how uncanny or AI-obvious the avatar appears, including lip-sync plausibility.
decisive for this rankingtransformation
This ranking is about whether the avatar convincingly plays the creator in the ad; if it looks uncanny or AI-obvious, the core product has failed. (3 of 3 judges)
What was given, what came back
Test input: FutureSmart AI – SaaS · text
Input — what we sent
The exact prompt
Product URL: https://aidemos.com/ Exact Script Hook: Finding the right AI tools shouldn't take hours. Script: I've been using FutureSmart AI to discover and compare AI tools in one place. It helps me find the right tool faster with real use cases, rankings, and detailed comparisons. If you regularly use AI tools for work, it's definitely worth checking out. Output Requirements - Vertical (9:16) - 20–30 seconds - UGC-style AI presenter - Natural AI voice - Captions enabled
A short-form UGC-style video ad prompt for a SaaS product, designed to test whether a tool can understand a software offering and turn a concise testimonial script into a believable ad.
Why this input is hard
- · SaaS product understanding
- · script quality
- · avatar realism
- · voice quality
- · caption quality
- · overall ad readiness
Output — unretouched
Also checked on this input — same tool, 2 other criteria
Script quality & persuasiveness⚠ StruggledThe delivery leaned corporate rather than creator-led, with limited emotional variation and no short-form optimization, which weakens UGC-style persuasiveness.Voice quality✓ WorkedThe voice was natural and easy to understand, with a clear, structured delivery.
Provenance
- Observation
- 1e3d92d7-08df-474f-b58d-7f8fbc3654ca
- Evidence run
- f91a51ec-b400-46fe-97f5-7e779fe17532
- Study
- Generate UGC-Style Video Ads With AI Avatars
- Research task
- 86baxegpg
- Tested at
- not recorded
- Source
- first-party
- Evidence state
- verified
- Proof shown
- input + output shown
- Cost / latency
- not captured
- Repeat run
- not captured
- Tester
- not captured
The last three rows are honest blanks, not placeholders — our capture has no field for them yet.
Query this
get_evidence({
tool: "synthesia"
})MCP · mcp.aidemos.com/api/mcp
Free with attribution.
Same input, same check — 5 other tools
measured on Avatar realism
AKOOL◐ MixedThe presenter stays consistent, but still reads as somewhat AI-generated and less natural in some moments.Creatify.ai✓ WorkedThe avatar presentation is visually convincing and close to a real human, with accurate lip sync, smooth delivery, and natural gestures.Topview AI⚠ StruggledThe presenter reads as AI-generated rather than human, with lip movements that are not always aligned to speech and minimal facial/body motion.VEED✓ WorkedThe avatar looked relatively polished and realistic for a basic UGC generator, and lip movements were generally well aligned with the generated voice.Vidnoz AI◐ MixedThe presenter is moderately realistic, but the avatar can still read as slightly AI-generated and less natural in some moments, even though lip movements are generally aligned.
This evidence is published in
From the same study (page rebuilt from a later run)
Real inputs and real outputs, no retouching · every cell queryable via API & MCP · aidemos.com