Stylised animation
A paper cut-out look has to survive five seconds of motion without sliding toward smooth 3D or photorealism.
2 models received different prompt text — show each
Each model is given the prompt its input contract calls for: a multi-reference model is told which reference is which character, a segmented model gets shot segments. The frame, the references and the requested length are identical throughout. This is the honest version of "the same prompt".
H3 Max, H3 ref2v, Wan 2.7, Seedance 2.0, Veo 3.1, LTX-2
The paper fox walks from left to right through the paper forest with a slightly jerky stop-motion gait, its ears twitching. Paper leaves drift down. Keep the paper cut-out look throughout: visible texture, cut edges, layered shadows. No smooth 3D shading, no photorealism. Gentle paper rustling, no music.
Kling v3 Pro
[{"prompt":"The paper fox walks from left to right through the paper forest with a slightly jerky stop-motion gait, its ears twitching. Paper leaves drift down. Keep the paper cut-out look throughout: visible texture, cut edges, layered shadows. No smooth 3D shading, no photorealism. Gentle paper rustling, no music.","duration":"5"}]Same character, separate call: The same fox, a different scene · how consistent each model was
| H3 Max | H3 ref2v | Kling v3 Pro | Wan 2.7 | Seedance 2.0 | Veo 3.1 | LTX-2 | Seedance 2.5 | |
|---|---|---|---|---|---|---|---|---|
| Lines heard | — | — | — | — | — | — | — | — |
| Voice | model's own | model's own | model's own | model's own | model's own | model's own | model's own | model's own |
| Motion | 0.64 | 2.47 | 3.69 | 2.11 | 1.00 | 7.50 | 10.96 | 7.29 |
| Scene changes | 0 | 0 | 0 | 1 | 0 | 0 | 0 | 0 |
| Loudness / peak | -31.4 / -7.1 | -38.4 / -19.5 | -52.9 / -27.6 | -34.8 / -15.7 | -46.5 / -27.6 | -15.3 / -5.1 | -46.3 / -22.1 | -42.7 / -24.1 |
| Length | 5.18s | 5.18s | 5.04s | 5.04s | 5.06s | 8.00s | 4.92s | 5.04s |
| Frame | 768×1344 | 768×1344 | 1072×1928 | 720×1280 | 720×1280 | 720×1280 | 576×1024 | 720×1280 |
| Wait | 7s | 4:06 | 3:34 | 2:07 | 2:22 | 35s | 55s | — |
| Attempts | 1 | 1 | 1 | 1 | 1 | 2 | 1 | 0 |
| Cost | $0.40 | $0.30 | $0.98 | $0.50 | $0.90 | $3.20 | $0.22 | — |
The read
Style held better than expected. H3 Max, H3 reference-to-video, Kling and Seedance all kept the layered paper look through five seconds of walking, with cut-edge leaves falling on cue; Kling's is the most consistent frame to frame. LTX kept the style but moved the camera instead of the fox, which is why its motion number is the highest on the row — a sliding background, not an animated character. Wan re-staged: a larger, redesigned fox walking toward camera, and by the last second a different fox with smooth shading, which is exactly the drift toward 3D the prompt said not to do.
One person's judgment from the takes above, 2026-09-03. The table is the evidence; this is the opinion.
what whisper heard
- H3 Max: (footsteps thudding)
- H3 ref2v: [SOUND]
- Kling v3 Pro: [BLANK_AUDIO]
- Wan 2.7: [SOUND]
- Veo 3.1: (soft music)
- Seedance 2.5: (keyboard keys clacking)