Text+Image-to-Vector Animation Generation on Lottie
0.9743CLIP ScoreLottieGPT-7B-Stage2
Evaluation results
| Method | Links | |||||||
|---|---|---|---|---|---|---|---|---|
| LottieGPT-7B-Stage2Input=Text+Image2026.04 | 0.9743 | 0.9886 | 0.0278 | 0.9472 | 0.9195 | 0.3029 | 97.83 | |
| Sora2Approach=Image2Video2026.04 | 0.8903 | 0.8661 | 0.3004 | 0.8182 | — | — | — | |
| KlingApproach=Image2Video2026.04 | 0.8661 | 0.8542 | 0.2783 | 0.6261 | — | — | — | |
| Claude Sonnet 4.5Mode=Few-shot (3-shot)2026.04 | 0.8499 | 0.9229 | 0.1273 | 0.7349 | 0.0043 | 0.3416 | 51.74 | |
| Veo3.1Approach=Image2Video2026.04 | 0.8487 | 0.5558 | 0.4417 | 0.6658 | — | — | — | |
| Gemini 2.5 ProMode=Few-shot (3-shot)2026.04 | 0.8476 | 0.8311 | 0.2136 | 0.698 | 0.0037 | 0.2862 | 32.17 | |
| Qwen3-235BMode=Few-shot (3-shot)2026.04 | 0.8162 | 0.9251 | 0.1721 | 0.7084 | 0.0028 | 0.2831 | 24.35 |