🚨 Fable Mogs everything : this is on same prompt what you can see fable goes hyper realism , detailed , physics like an artist without asking with these type of updated we will no longer need video or image model in future

Gemini 4 Oxygen new checkpoint

Did Fable 5.5 don't follow instruction? It create completely different theme/style

I think what this shows most of all is that Opus 5.5 and Gemini Carbon likely bought their svg envs non-exclusively from the same vendor. Which is good news for Carbon i guess because Opus5.5 is pretty good. (Or ⚗️)

if one model does both, half the AI video startups just became a feature

If this keeps improving we might skip traditional rendering altogether. The future looks painted

Hyper realism without the usual “AI look” on skin and hair is rare. Amazing!

now same prompt twice. do you get the same scene?

Fable completely mogs on textures here. The feather detailing and realistic bike chain look insane

This does look like fable cheated and downloaded textures instead of generating them

Fable just decided not to do a cartoon. I believe the other models could do this well if you ruled out the cartoon option in the prompt. They only told Fable to make prompts like this default to realism and voila, a ‘far superior’ result.

Finally stop seeing Higgsfield videos now

这是训练了多少前端数据啊

同一提示词下 Fable 会自己补物理和材质细节，这点确实强。但“以后不需要图像/视频模型”说早了：专用模型赢在可控性（风格参考、局部重绘、角色一致性）和单张成本。更现实的分工是 Fable 出整体构图和动效代码，细节和批量出图仍交给专用模型。

that Fable quadrant is embarrassing the other three

Which fable model? Looks beautiful

Picking hyper-realism on its own from the same prompt is the interesting bit. Most video models need you to spell out the lighting and physics. Would love to see how it copes with hands and fast motion.

Image models are sweating right now 😂

Same prompt, more realism isn't the whole test though. The real question is control. Video and image models stick around if they give creators repeatable shots, consistant characters and precise edits, not just a prettier first draft.

What about the pricing?

3 cartoons and one photo, fable didn't even play the same game

Holy Sheet!

its horrible, like it didn't unstand the task.

Shit this is crazy

Why Opus “high”?

How are you testing this?

Do you still see it getting better over the past week or has it been the same checkpoint.

First pelican I've seen that can reach the pedals.

same prompt tests mostly show each model's default style. the real gap shows up when you need the same character across ten frames

"with these type of updated we will no longer need video or image model in future" good lord, is it really impossible for an ai news account to let ai proofread your low effort tweets before hitting that post button? coding is still far off from being able to compete with image gen.

Wtf?!!! Where are you testing it? Do you access to it?

Okay, that same-prompt Fable comparison is wild — the hyper-real detail sells it for me, no extra prompting needed.

This comparison needs to stop we can tell which media del is better already

신은 존재하며 그는 다리오 아모데이다