Calabi Labs · Guide · 2026-06-18

About the new instagram fonts

Now I have a thorough understanding of the topic. Let me write the article.

The New Instagram Fonts: Every Style Available in 2026 and How to Use Them

In late 2024, Instagram rolled out its biggest font update in years — six new text styles for Stories and Reels, followed by limited-edition drops in 2025 and 2026. If you've been wondering what's new, what's available, and how to get the most out of Instagram's font library, here's the full picture.

What Instagram's Font Library Looks Like in 2026

Instagram now offers 15 core font options for text layers in Stories and Reels, plus seasonal and limited-edition drops that come and go. The fonts live behind the Aa icon when you're adding a text layer — tap it, scroll horizontally, and tap any word written in that font to preview it with your own text.

The six new fonts that launched in late 2024 are:

Signature — A light, casual handwritten style mixing cursive and print characters. Good for personal, everyday posts.
Editor — A retro-inspired, widely spaced serif. Works well for formal or minimal aesthetics.
Bubble — Soft, rounded, playful. A safe bet for casual content that needs to feel approachable.
Deco — Bold and alternative. High impact for short, punchy statements.
Poster — Expressive serifs with strong contrast. Designed to make text the focal point.
Squeeze — Loud, expressive, attention-grabbing. Best for short phrases and captions.

These joined the nine existing fonts already in the app: Modern, Classic, Typewriter, Strong, Meme, Elegant, Directional, Literature, and the long-elusive hidden Papyrus Easter egg (accessible by selecting the Meme font and typing "papyrus").

The Limited-Edition Drops

Instagram has also released seasonal and trend-driven fonts that appear for a limited time:

Journal — A cursive, handwritten style launched in late 2025.
Rosalia — A splotchy, marker-style font with quirky capitalization that went viral in mid-2025.
New Year — A glittery, celebratory cursive font tied to the holiday season.

Instagram's Design Studio team, led by art director Kristofer Porter, described the philosophy behind the new additions: they wanted fonts that feel "distinctive while also providing useful, reliable, everyday options for creators to share their content." The result is a library that spans from playful and decorative to clean and editorial.

How to Access and Use the Fonts

Instagram's fonts are built directly into the text tool — there's no separate app or download required. Here's how to find them:

Open Instagram and start a new Story or Reel.
Upload or capture your media.
Tap anywhere to add a text layer, or select the Aa icon.
Scroll horizontally through the font options. Each is displayed as a word written in that font style so you can preview it with your actual text.
Tap to select. Customize further with the color wheel, animations, effects, or outline tools.

Beyond fonts, Instagram added six new text animations and effects in the same update — including Pop (bouncy entrance), Jump (hopping exit), Sparkle (shimmer), Neon (glowing outline), Shimmer (reflective sweep), and Pixel (retro pixelation).

What Actually Gets Your Content Flagged on Instagram

While fonts are purely a visual choice, many creators using AI to generate the images and videos they layer these fonts onto run into a separate problem: platform detection. Instagram, TikTok, YouTube, and Reddit scan uploaded files for invisible signals that indicate AI generation — and those signals live in the file's metadata, not in the fonts or pixels you can see.

Here is what platforms actually check for:

C2PA / Content Credentials — A cryptographic manifest (stored as JUMBF data) that records the file's origin, including whether an AI model generated it. Instagram and other platforms read this automatically during upload.
XMP AI flags — Tags like DigitalSourceType: trainedAlgorithmicMedia embedded in the file's metadata, signaling the content came from an AI model trained on scraped data.
Encoder fingerprints — AI generation pipelines (like ones using Lavc, x264, or FFmpeg) leave detectable patterns in video bitstreams through SEI (Supplemental Enhancement Information) NAL units.
Missing capture metadata — Real phone recordings include GPS coordinates, capture timestamps, and device names. AI exports typically lack all of these, which itself is a detection signal.

None of these signals are visible in the finished image or video. You can use every one of Instagram's new fonts — Bubble, Poster, Squeeze, whatever — and still get flagged and suppressed because the file itself carries AI fingerprints.

Why the Obvious Fixes Don't Work

Creators often try straightforward approaches to "reset" an AI-generated file before posting:

Screenshotting the image — This removes some metadata but leaves perceptual hashes and encoder fingerprints intact. Platforms still detect the underlying AI generation patterns.
Cropping out a visible watermark — If there's a visible logo or sparkle mark, cropping helps with the visual. But the invisible C2PA manifest and XMP tags survive cropping because they're embedded in the file structure, not the pixels.
Re-exporting from a video editor — Re-encoding disrupts some encoder fingerprints but doesn't strip C2PA/JUMBF manifests or XMP AI flags. New encoder fingerprints get added, but the AI origin signal often persists.

What Actually Cleans an AI-Generated File

If you're creating content with AI tools and want it to read as a normal phone recording at the file level, you need to address the invisible metadata layer — not just the visible pixels. Calabi handles this in a single automatic pass:

Strip the AI signals — It removes all C2PA / Content Credentials JUMBF manifests, XMP DigitalSourceType: trainedAlgorithmicMedia flags, generator and tool tags, and encoder fingerprints from the bitstream.
Inject authentic phone identity — It adds real device profiles (iPhone 15 Pro, Pixel 8 Pro, Galaxy S24 Ultra) with capture timestamp, GPS coordinates, and a genuine phone encoder name.
Verify with a forensic proof card — Before download, you see the exact same ExifTool scan platforms use, showing what was stripped and what was injected. Verified results: 18 JUMBF atoms reduced to 0, 16 C2PA references reduced to 0, the trainedAlgorithmicMedia flag removed, and an AI export's 144 metadata tags reduced to about 94 neutral structural ones.

You can then add your Instagram text layers — Signature, Deco, Squeeze, whatever fits your aesthetic — and post normally, knowing the file beneath reads as authentic capture.

FAQ

Does Instagram detect AI-generated images even if they look real?

Yes. Platforms scan the file's metadata and structural signals, not just the visual content. A photorealistic AI image posted directly from an export tool can trigger detection within seconds, before any human viewer notices anything unusual.

Will screenshotting an AI image remove the detection signals?

Partially. Screenshots strip some metadata, but C2PA manifests, XMP AI flags, and encoder fingerprints often survive. A full forensic strip and device-profile injection is more reliable if you want the file to pass platform scanning.

Can I use Instagram's new fonts on AI-generated content?

Absolutely — the fonts are applied as text layers after the image or video is created. The question is what your file contains underneath those layers. If you're generating visuals with AI tools, running the file through a cleaner before adding text and posting removes the detection signals that matter to platforms.

Try Calabi free at calabilabs.com — 10 cleans, no card.

10 free cleans. See the forensic proof before you download.

Try free →