Head to head: Bernini-R Edit Video vs Luma Ray 3.2 Image to Video
Bernini-R Edit Video vs Luma Ray 3.2 Image to Video
This one wasn’t subtle. Across four judged prompts, Bernini-R Edit Video consistently delivered tighter prompt adherence, cleaner motion control, and more reliable scene logic than Luma Ray 3.2 Image to Video.
Bernini-R Edit Video wins this matchup decisively: **34.0 to 26.1 overall, with 97% confidence**, and a clean **4–0 sweep in task wins**. That matters, because this wasn’t a case of one flashy clip carrying the result. Bernini-R was better across the board on the things that actually decide whether a video model is usable: following the prompt, controlling camera motion, and keeping scene behavior coherent from start to finish. The clearest separation showed up in prompt fidelity. In **Stormglass Marsh Drift**, Bernini-R gave the judges the shot that was actually asked for: a single yellow skiff, blue-hour post-rain mood, red poncho at the stern, steam over the marsh, and a stable drifting feel. Luma’s version had decent water motion, but it drifted off brief with brighter daylight energy, apparent occupants, and a more chase-like perspective. The same pattern held in **Lighting transition**, where Bernini-R nailed the locked-off living-room setup and the lamp visibly coming on during the dusk-to-cool-blue shift; Luma’s lamp behavior read backward, dimming or switching off instead of delivering the requested flicker-on beat. Bernini-R also looked more controlled when the prompt demanded camera discipline. In **Camera motion control**, it produced a recognizable smooth orbit around the ramen bowl at a stable height, with stronger food detail and a more convincing tabletop scene. Luma had passable ambience and steam, but the move read like a lateral drift rather than the requested 180-degree orbit, and the bowl itself looked less specific and less appetizing. That’s a recurring issue here: Luma often gets the vibe, while Bernini-R gets the shot. Even the closest category, **Hail Cart Behind the Salt Shed**, still ends up in Bernini-R’s column in the aggregate result. The judge notes show some disagreement on which model handled the occlusion beat more faithfully, but that split only reinforces the broader point: when there was any ambiguity, Bernini-R’s stronger clarity, continuity, and overall compositional reliability kept it ahead across the full evaluation. Luma showed flashes of atmosphere, but atmosphere alone doesn’t win head-to-heads. **Final call: Bernini-R Edit Video is the better video model here, and not by a hair. It was more faithful, more controllable, and more cinematically coherent in every judged task that counted.**
Stormglass Marsh Drift
At blue-hour after a summer squall, a single continuous 16:9 shot glides low on a fluid gimbal along the edge of Stormglass Marsh as a yellow skiff with a red rain poncho tied to its stern drifts diagonally through frame; the camera slowly tracks right and slightly forward while cattails bend in uneven gusts, silver ripples spread across black water, torn lavender clouds race overhead, and thin steam lifts from warmer mud banks into the cool air, all moving naturally and continuously in moody post-rain light.
Model A matches the prompt much more closely with a single yellow skiff drifting through a blue-hour marsh, a red poncho tied at the stern, moody post-rain lighting, steam over the water, and stable continuous camera motion. Model B has decent marsh motion and wake detail, but it breaks prompt adherence with brighter daylight lighting, apparent occupants in the boat, and a less convincing match to the specified drifting gimbal shot and atmosphere. (Second judge pass, order swapped — scores are the average of both: Model A matches the prompt more closely with blue-hour post-rain mood, a yellow skiff drifting diagonally with a red poncho at the stern, visible marsh steam, and stronger overall composition; its frames also look cleaner and more cinematic. Model B has decent continuity and water motion, but it reads more like a chase shot behind a boat with an occupant, misses the specified drifting feel and moody lighting, and is softer and less faithful overall.)
Lighting transition
A living room at dusk as the sun sets over a few seconds: the warm light visibly dims and shifts to cool blue while a lamp flickers on, the transition smooth and believable, locked-off frame, 16:9.
Model A better matches the prompt with a believable dusk-to-cool-blue transition in a locked-off living room and a lamp that visibly comes on, while maintaining strong visual quality. Model B has a nice moody atmosphere, but its lighting progression is less aligned with the prompt because the lamp appears to dim or turn off rather than flicker on, making the transition feel less correct. (Second judge pass, order swapped — scores are the average of both: Model A better matches the prompt by showing a believable dusk-to-cool-blue transition with the lamp turning on in a smooth, locked-off living-room shot, while maintaining stronger clarity and composition. Model B has a nice color shift, but the lamp behavior appears inconsistent with the requested flicker-on transition and the scene becomes overly dark with weaker visual fidelity.)
Camera motion control
A single smooth orbit (roughly 180 degrees) around a steaming bowl of ramen on a wooden table, the camera circling at a constant speed and height in one continuous move, warm restaurant light, 16:9.
Model A better matches the prompt with a clearer smooth orbit around a steaming ramen bowl at a fairly constant height, and it maintains stronger detail and appetizing presentation. Model B has warm restaurant ambience and decent consistency, but the camera move reads more like a mild lateral drift than a pronounced 180-degree orbit, and the bowl lacks ramen toppings/detail and overall sharpness. (Second judge pass, order swapped — scores are the average of both: Model A better matches the prompt with a clearer, smoother orbit around a steaming ramen bowl at a stable height, while maintaining stronger detail and coherence in the bowl and table setting. Model B has warm restaurant ambience and steam, but the motion reads more like a mild lateral drift than a pronounced 180-degree orbit, and the ramen itself is less visually convincing.)
Hail Cart Behind the Salt Shed
In cold late-afternoon sunlight under a bruised green storm sky, a single continuous 16:9 shot dollies left across a windy gravel lot at the remote Brine Nine weather outpost as a small orange utility cart hauling two blue hail-sample barrels drives steadily from right to left, disappears completely behind a corrugated salt shed for a full beat, then re-emerges on the far side unchanged with the same barrels, speed, and orientation; loose warning tape flutters, dust skims the ground, and distant sheets of rain drift across the horizon, creating a tense, expectant mood.
Model A matches the key action most cleanly: the orange cart with two blue barrels moves right-to-left, fully disappears behind the corrugated shed for a beat, and re-emerges unchanged under a dramatic green storm sky with convincing dust and tension. Model B has solid atmosphere and camera motion, but the cart appears to approach the shed from the wrong side relative to the prompt and the occlusion/re-emergence reads less precisely, reducing adherence and consistency. (Second judge pass, order swapped — scores are the average of both: Model B follows the key action more faithfully: the orange cart with two blue barrels moves right-to-left, disappears fully behind the corrugated shed for a beat, and re-emerges unchanged while the windy, stormy atmosphere remains coherent. Model A has strong visual clarity and dust effects, but the framing appears more static than a left-dollying shot and the cart’s re-emergence/continuity reads less cleanly, with added on-screen text hurting prompt fidelity.)
Matchup powered by OpenRouter.