I'm a landscape photographer. I batch
Pre-upload checklist for Topaz Photo AI & Instagram AI Labels
- Finalize your export — no extra apps after cleaning.
- Spot-check one hero image in the AI metadata checker.
- Strip metadata with Remove AI Label — 30 images per batch.
- Upload before posting to Instagram, Facebook, or Pinterest.
- Deliver a Social_Ready folder so clients never re-upload RAW files with C2PA.
One master JPG is enough
Keep a master JPG after Lightroom or Photoshop. Remove C2PA and XMP once, then reuse for feed, Story, ads, and marketplace listings — as long as you do not send the file through Canva or mobile AI apps again. Each extra app can re-attach provenance markers.
Common mistakes with Topaz Photo AI & Instagram AI Labels
- Mixed carousel slides — half cleaned, half not; AI Info returns on the next flagged frame.
- Re-export after cleaning — Canva and Adobe Express re-attach provenance.
- Screenshots instead of exports — do not reliably fix metadata.
- Fixing live posts — Instagram does not strip C2PA from stored files; export the original, clean, republish.
Cross-posting and live posts
Same JPG for Instagram, Pinterest, and TikTok? Remove metadata once before every channel. Meta and Pinterest scan C2PA and XMP. To fix a live post, download your original export, clean in the browser, upload again — see Remove AI Info from Instagram.
Reduce support tickets
Email clients: "If you see AI Info, it is almost always edit metadata — use Social_Ready." Link AI label false positives in onboarding PDFs.
EXIF vs C2PA
Need camera EXIF for archive or print? Strip only C2PA and XMP, keep standard EXIF when your workflow allows. The checker shows which blocks are present before you clean.
Workflow summary
Inspect one file → batch-clean with Remove AI Label → upload cleaned JPG → deliver Social_Ready copies. Browser-based processing keeps files on your device — useful for client galleries and listing photos.
*Use on files you own. Follow platform disclosure rules where they apply — see our [disclaimer]
Why Topaz triggers AI Info
Topaz products apply machine-learning models for:
- Upscale (Gigapixel / Photo AI)
- Denoise and Sharpen
- Face recovery on portraits
Exports may embed C2PA or XMP marking AI-assisted processing — similar to Lightroom AI Denoise, but from a different vendor stack.
| Workflow | Label risk |
|---|---|
| RAW → Lightroom → Topaz upscale → Instagram | High |
| Topaz only on JPG export | High |
| Manual export, no AI tools | Lower |
Photographer delivery workflow
- Process in Topaz Photo AI / Gigapixel.
- Do not upload directly to client Dropbox for social use.
- Metadata checker on one frame per batch job.
- Remove AI Label — batch 30 images per pass for wedding or landscape sets.
- Deliver
Social_Cleanfolder alongside print masters.
Tell clients: "Instagram-ready folder has metadata cleaned; print folder untouched."
Stock and print sellers
Some marketplaces scan metadata. Cleaning before Instagram marketing and stock upload keeps one consistent master.
Topaz + Adobe round-trips
Lightroom → Topaz → Photoshop can stack markers from multiple apps. Clean after the last save, not after Topaz alone if Photoshop runs afterward.
Gigapixel batch jobs, print masters vs. social exports, and version notes
My Gigapixel queue runs overnight — four hundred archive frames from a decade of trail photography, upscaled for a gallery show and a simultaneous Instagram retrospective series. The batch finished clean visually. The metadata did not: Topaz Photo AI 4.x builds write C2PA differently than Gigapixel AI 7.x depending on which model handled denoise vs. upscale. I learned this after three client-facing exports in one week — two labeled on Instagram, one not — all from the same overnight batch preset.
Gigapixel batch processing treats metadata as an afterthought because photographers think in pixels. You queue two hundred wedding reception frames, apply Recover Faces and Denoise, wake up to a folder of gorgeous 60-megapixel JPGs, and ship a Social subfolder to the couple without a checker pass. Every file in that folder may carry identical provenance blocks — one preset, one metadata signature, two hundred Instagram posts waiting to fail if the family shares broadly during thank-you season.
| Output destination | Typical resolution | Metadata priority | Clean? |
|---|---|---|---|
| Print master (Giclée, lab) | Full upscale, 16-bit TIFF/JPG | Archival — some labs ignore C2PA | Optional for print-only |
| Client social folder | 2048–4096px long edge | High — Meta scans at upload | Required |
| Instagram photographer portfolio | Web JPG | High | Required |
| Stock agency submission | Varies by agency | Some agencies flag AI metadata | Check agency rules |
Print vs. social split saves confusion in client delivery emails. Deliver Print_Archive/ untouched for wall art and albums. Deliver Social_Clean/ with C2PA and XMP stripped for Instagram, Facebook, and announcement posts. Never assume clients understand the difference — spell it out: "Post from Social_Clean only; Print_Archive is for framing." Wedding and landscape clients routinely grab the highest-resolution file in the zip; that is often the uncleaned upscale.
Version notes matter because Topaz updates change export behavior mid-season. After a Photo AI or Gigapixel point release, re-run the metadata checker on one export before batching another three hundred files. Community reports from 2025–2026 note C2PA embedding toggles and model-specific provenance strings that did not appear in earlier builds. Your preset from March may not match your preset from June even when sliders look identical.
For hybrid Photo AI + Gigapixel chains — denoise in one, upscale in the other — clean once after the final save. Intermediate Topaz exports can stack markers with Lightroom or Photoshop steps downstream. Build checker → clean as the last line in your export automation script, not an optional step you remember when a client complains.
Gigapixel-specific batch naming: Label output folders with Topaz version numbers (Giga_7.2_batch042) so you can trace which builds produced labeled exports if a client report surfaces weeks later. Version mismatches between your workstation and a second editor's laptop are a common source of "it worked last month" confusion.
Print lab handoff: If your giclée lab requests untouched files, deliver print masters separately and never assume the lab will strip C2PA before the client posts a phone photo of the framed print — that secondary photo is usually clean, but the original file they share from email might not be.
Keep a one-line README in each client delivery zip explaining which folder is for Instagram — confusion drives most false-positive support tickets.
Full upload prep for Instagram promos: Remove AI Info from Instagram.
Related reading
- Lightroom AI Denoise & Instagram
- Wedding photos flagged as AI
- What is C2PA?
- AI label false positives
upscale and denoise with Topaz Photo AI before printing and Instagram posting. The scenes are real — hiking trails, coastlines, no generative AI anywhere in the pipeline.
After a firmware-style update cycle in 2025–2026, clients started reporting AI Info on exports I delivered for their Instagram feeds. The checker showed C2PA on files that never touched Midjourney or Firefly.
Topaz is doing legitimate enhancement, not synthetic generation. Platforms still read provenance metadata and apply disclosure — another false positive for photographers who treat Topaz like a sharper export button.
See disclaimer.
See disclaimer.
Post Topaz-edited photos without Instagram AI Info
Export from Topaz, remove C2PA and XMP, deliver or upload cleaned files.
- Export from Topaz Photo AI or Gigapixel — Save JPG or PNG at your delivery resolution after upscale, denoise, or sharpen processing.
- Inspect the export — Run the file through the AI metadata checker — Topaz exports frequently show C2PA blocks.
- Strip C2PA and XMP — Clean in the browser without re-processing pixels in Topaz.
- Upload or deliver — Use cleaned files for Instagram, client galleries, and stock submissions when metadata labels are unwanted.
Frequently Asked Questions
Does Topaz Photo AI add metadata that triggers Instagram AI Info?
Topaz exports often include C2PA or XMP tied to AI upscaling, denoise, and sharpen models — Meta may label the upload even when the source photo was shot on a real camera.
Will removing metadata undo Topaz upscaling quality?
No. Cleaning removes invisible provenance tags only; upscaled pixels and sharpening remain unchanged.
How to remove AI info from Topaz exports before Instagram?
Save JPG or PNG from Topaz, run the metadata checker, strip C2PA and XMP in the browser, then upload the cleaned file.
Should I clean after every Topaz batch export?
Yes — build clean-into-export as the last step before client delivery or social upload, especially for wedding and portrait batches.
Is Topaz Gigapixel different from Photo AI for metadata?
Both product families may write provenance on export depending on version and settings — inspect the actual file you post, not assumptions by product name.