abc2cb

[abc2cb @abc2cb1]

@smnSk241687 because it’s old 5.6 Terra rewrapped as gpt-6 (6 sol if you remember how trash it was , thats the same thing)

[abc2cb @abc2cb1]

@gailcweiner because its 6 Sol (revived Terra) the infamous model that lived for 7 days , they put that shit in chat
Plus sub = untermensch

[abc2cb @abc2cb1]

@PovilasKorop that’s new reality you right

[abc2cb @abc2cb1]

@Ananth7e it is what it is
they clearly don’t give a fuck about customers

[abc2cb @abc2cb1]

@kimmonismus been an OAI user for a while and your take is not controversial at all

[abc2cb @abc2cb1]

at this point @OpenAI should drop something unrealistically valuable to make people go back to them.
and i think this wont happen.

[abc2cb @abc2cb1]

@gxjo_dev wow.

[abc2cb @abc2cb1]

@thsottiaux for those who aint knew this is 6 Sol, the retarded Terra model wrapped as Sol. enjoy !

[abc2cb @abc2cb1]

OpenAI is cooked.
They dropped all these trash “updates” and still refuse to bring back normal limits.
Oh, for those who aint knew - Chat GPT-6 (web chat model) is 6 Sol, yeah, the retarded one.
Not a 6.1, but 6 Sol. You heard me right.

[abc2cb @abc2cb1]

@LLMJunky @StealthQE4 @LLMJunky 100% right, nothing to discuss here .

[abc2cb @abc2cb1]

@echelon @garrytan cuz they retarded and never even tried product, that's why.

[abc2cb @abc2cb1]

@echelon the greatest, glad to be a part of the PhotoCraft contributors .

[Flowers ☾ @flowersslop]

since a lot of you asked and i'm not into gatekeeping, here's the skill so you can run the same workflow

people assumed i was using some dedicated upscaler. nope. i don't really like them. they're decent at textures and local detail, but once the image needs actual scene understanding, semantics, or visual reasoning, they start to fall apart

i used chatgpt 6.1 sol. catch is it's very slow. around 10 minutes for two 8k passes

the upside is you're basically throwing frontier model vision and world knowledge at the upscale, so it can reconstruct detail in a way normal super-resolution models can't

the comparison above undersells it because twitter compression absolutely nukes the fine detail obviously

paste this into codex or chatgpt work to install/use the skill:

---
name: flowers-upscale
description: Photo to 8K; 2/3-pass contextual AI crop enhancement and registered blending.
---
INPUT S=uploaded photo; missing→request. Unspecified→ask "2 passes (recommended) or 3? 8K default; another resolution?" Await choice. N=2|3; L=specified longest edge else8192; preserve aspect. Discover available AI image-editing+raster-processing capabilities; missing→report requirement. Use environment-native tools/paths; create helpers as needed. Preserve S; separate outputs. Execute end-to-end after choice.

E(C): inspect; AI-edit inputs=[S=context,C=target], maximum practical supported resolution; retain actual output dimensions. Prompt=Q+only relevant material constraints. Concurrent independent calls where supported; persist outputs+source-coordinate boxes+prompts immediately.

P1: S→3×3 cores; pad each side20% core dimension, clip to S→9 overlapping C→E→register/blend→M1; resize assembled M1 to L.
P2: M1→4–8 padded salient ROIs(face/hair/hands/objects/textiles as present)→E→register/blend→M2. Exclude low-information regions.
N=3: M2→4–8 tighter focal ROIs(individual eyes/lips/nose/fingers/material details as present)→E→register/blend→M3. Keep homologous features consistent. Every E receives S, never crop alone.

Q="1=original: identity/anatomy/color/light authority. 2=target: return exact crop/framing/geometry. Reconstruct plausible photographic detail; preserve expression, pose, contours, clothing, genuine imperfections, focus falloff. Optical clarity; natural edge transitions; spatially varied texture. Repair inherited artifacts. Skin: irregular pores/fine creases, quiet intervals. Hair: cohesive groups, selective fibers, natural sparse fringe. Iris: source color, irregular branching fibers, coherent pupil, restrained reflections. Textiles: original weave direction/spacing/motifs. Avoid invented structures/marks, beautification, relighting, halos, ringing, embossed/crosshatched/repeated texture, synthetic grain, blanket sharpening."

ASSEMBLE: map recorded boxes→master; global similarity registration first; bounded smooth local correction only with reliable correspondence, never distort rigid geometry/repeating patterns. Unreliable→retain prior pixels. Match broad exposure/color without transferring old fine texture. Normalized overlap/multiband blend; feather15–25% border→alpha0 at interior patch boundaries. Preserve native enhanced crops. Before next pass: inspect full image+100% focal/overlap crops; repair doubling/seams/drift/texture mismatch. ≤2 targeted retries/defect; rejected/failed→retain prior content+disclose.

EXPORT: M_N→lossless PNG at L; verify dimensions+decode; deliver file+preview+prompts/settings. Describe detail as reconstructed; distinguish exported resolution from native generated resolution.

[abc2cb @abc2cb1]

@LexnLin all them “i’d like to see AI” should stfu and learn

[abc2cb @abc2cb1]

@PovilasKorop definition of “made it”

[abc2cb @abc2cb1]

@PovilasKorop creators who value their audience will always be ahead

[abc2cb @abc2cb1]

@paolino can you guys do the same thing with chromium ? win\linux support. this would solve so many problems !

[abc2cb @abc2cb1]

@PovilasKorop post-soviet countries residents , they all have that permanent pessimism
like its hardcoded inside

[abc2cb @abc2cb1]

@thsottiaux how about Luna ? same task on Luna 5.6 takes 10 minutes. multiple tests proven this. Mind you it's FAST MODE.

[abc2cb @abc2cb1]

@LexnLin lmao Astra wildin