Author archive
Creatos Editorial
AI news and analysis published by Creatos. Articles are researched and written with AI assistance, with sources linked and observations distinguished from reported claims.
Published entries
What is an infinite canvas? A file and handoff checklist
Choose an infinite canvas by checking file imports, editable projects and export limits. Includes a three-file evaluation checklist and a handoff folder template.
Claude image review: keep small text readable
Compare a contact sheet, separate images and detail crops using Claude image limits, with reproducible pixel and token calculations for four banner variants.
Holo4 click coordinates: fix crop and Retina offsets
Map Holo4 normalized coordinates from a cropped screenshot back to a design app. Includes a worked Retina example and locally tested Python conversion.
Save Colab images before a runtime reset: verify the copy
Keep generated PNGs and settings outside the Colab runtime. Use a locally tested copy-and-hash helper, then check the files independently in Drive.
MiMo video review: why short details disappear
A worked sampling example shows why MiMo can miss a brief label, how fps differs from resolution, and what to check before approving a product video.
Build a useful Claude brand-review skill before packaging a plugin
A proposed four-case test for a Claude brand-review skill: correct artwork, wrong wording, missing references and optional style preferences.
Opus 5.5 image review: when prompt caching saves money
Calculate Opus 5.5 image-review costs with cache writes, five-minute and one-hour lifetimes, and a twelve-request example you can recalculate.
ChatGPT image privacy: before you upload client images
Check ChatGPT image training, retention and upload permissions before sharing client work, with an example showing when removing a logo is not enough.
GitHub Copilot canvas: plan an image review board
A proposed Copilot canvas brief for image reviews: bind approval to a file revision, separate recommendations, and check stale previews and saved decisions.
Can LongCat 2.5 generate images? Choose the right model
Separate LongCat 2.5 image understanding from LongCat-Image generation and Image-Edit, with a proposed poster correction brief and concrete output checks.
When does Seedance 2.5 Draft mode save money?
Calculate Seedance 2.5 Draft mode costs when keeping several video variants. Includes a break-even formula, six-candidate example and visual shortlist workflow.
Gemini 3.8 TTS voice consistency: reuse the narrator
Keep a narrator consistent when replacing voiceover lines. Use Gemini 3.8 TTS voice IDs, separate delivery cues and review each edit against approved audio.
AI project handoff template: decisions, assets and next steps
Resume creative work with a portable AI project handoff. Separate approved decisions from temporary status, preserve asset versions and resolve conflicts before generating.
Turn voice feedback into a clear design revision log
Convert spoken design feedback into specific changes with an object, constraint and review test. Includes a poster example, conflict handling and a copyable prompt.
AI subtitle translation: preserve SRT timing and product terms
Translate subtitles with a glossary, fixed SRT timestamps and a separate reading-speed check. Includes an English-to-French example and a copyable brief.
AI video character consistency: a three-shot continuity sheet
Plan consistent AI video characters with reference images, a three-shot worksheet and a cut-point review. Includes a fictional courier sequence and repair decisions.
AI presentation source map: turn research into five useful slides
A five-slide source map, worked example and copyable prompt for AI presentations. Keep claims, evidence and missing facts visible before designing the deck.
Screenshot to website with AI: what to include in the handoff
Prepare screenshots, exact copy, assets and mobile behavior for an AI website builder, with a worked portfolio brief and a clear review process.
Can DeepSeek read images? V4.1 Flash support explained
Which DeepSeek model accepts images, how the 600-image API limit interacts with file size, and why legacy aliases cannot compare old and new models.
Suno v6 migration: what to save for a video project
Preserve the exact Suno audio, export ranges and cue notes used in your video edit. Understand what changes with v6 and check your saved files.
How to compare AI answers for a creative brief
Compare AI proposals against one creative brief. A worked storyboard example shows what to keep, what to reject, and how to record the final decision.
ChatGPT Images 2.5: how to review an image edit
Review ChatGPT Images 2.5 edits against the source and your last approved version, with a concrete poster example and a reusable checklist.
AI agent productivity: what 3.1 workdays actually measures
Understand OpenAI’s 3.1 agent-workdays metric, then measure your own AI workflow with accepted outputs, review time and a worked example.
Blender AI workflows: MCP, computer use, or Python?
Choose a Blender AI workflow by the task: scene inspection with MCP, interface work with computer use, or repeatable Python jobs. Includes output and retry checks.
GPT-6 Astra prompting: three task briefs to try
Three GPT-6 Astra prompt examples for model demos, script editing, and website work, with an annotated edit and checks against the source material.