We ran two I2V scenes—a slow dolly-forward through an office corridor and a static time-lapse-like sunrise over mountains—across four video models. Veo 3.1 Lite delivered the best cost-performance at $0.12 per 4 seconds and won both prompts on average. Veo 3.1 and Seedance 2.0 also produced high-quality results for the static landscape but were much more expensive. Kling 3.0 V3 Standard failed both scenes.
Results
| Model | Avg Score | Wins | Cost / 4s | Verdict |
|---|---|---|---|---|
| Veo 3.1 Lite | 3.84 | 2 | $0.1200 | Best cost-performance |
| Veo 3.1 | 3.75 | 1 | $0.8000 | High quality, expensive |
| Seedance 2.0 | 3.84 | 1 | $1.2096 | High quality, expensive |
| Kling 3.0 V3 Standard | 2.38 | 0 | $0.3200 | Not usable in these scenes |
Key takeaways
- Veo 3.1 Lite is the clear winner for real estate and landscape I2V at $0.12 per 4 seconds.
- Static, background-focused scenes such as a mountain sunrise are far easier for all models than moving-camera shots.
- Seedance 2.0 matches Veo 3.1 Lite in quality but costs about 10x more, making it better for single high-value deliverables.
- Kling 3.0 V3 Standard produced jittery, unusable results in both real estate scenes tested.
Verdict
For real estate and landscape I2V, Veo 3.1 Lite is the best default for volume and cost; Seedance 2.0 is an option for premium one-offs; Kling 3.0 V3 Standard is not recommended based on this limited test.
Note
This is an English summary of the full Japanese benchmark article. Methodology, prompts, all outputs and cost details are documented in the original post.
Read the full article in Japanese →