Image Compression & Utility Presets
File-size limits show up in email attachments, web upload forms, document submission portals, and chat platforms. Email caps run around 25 MB total but practical limits are tighter — many corporate gateways reject single attachments over 5 MB. Government portals and visa applications often cap individual image uploads at 100-500 KB. Each compression preset here binary-searches JPEG quality to find the highest setting that stays below the target size.
An image that won't upload is an image that won't be processed. Document portals, ticket systems, and email gateways enforce size limits hard — exceeding the cap means resubmission. Compressing locally to a known target size removes the guesswork: the file fits the upload form on the first attempt, with the maximum quality the size budget allows.
How to choose the right preset
- Pick the file-size cap that matches your upload form. 100 KB for tight document portals, 1 MB for most web forms, 5 MB for email attachments. The card subtitle shows the target size.
- Smaller cap = lower quality at the same dimensions. Compressing a 12-megapixel photo below 100 KB will visibly soften — pair the cap with a maxWidth ceiling (the presets here do this automatically).
- Compress as the last step. Each lossy save round-trip discards detail. Do all your edits first (crop, colour correction, sharpening), then run the compression preset as the final step.
- Strip metadata when sharing publicly. EXIF can leak GPS coordinates, camera serial, and the photographer's identity. The strip-metadata preset removes these without touching the image content.
- Plan around the recipient's connection. Sending over slow mobile data, satellite, or in-flight Wi-Fi? Compressing aggressively makes the upload finish before the connection drops.
Every Utility & Compression preset
All 12 recipes in this category. Click any card to land in chat with the recipe pre-loaded — drop an image to run.
Common mistakes
- Setting the target lower than the source can sustain — a 12-megapixel photo squeezed below 100 KB will visibly soften. Pair the limit with a maxWidth ceiling.
- Compressing then editing — each save round-trip through JPEG compounds artefacts. Run compression as the final step.
- Using a generic PDF compressor for image-heavy PDFs — if the original is a photo, compressing the image directly is faster and produces a smaller file than wrapping the result in a PDF.
- Compressing already-compressed files — running this preset on a JPEG that's already at the cap doesn't make it smaller; the preset finds the highest quality that fits the budget.
- Stripping metadata from images you might need EXIF for later — the strip-metadata preset is one-way; keep the original if you might need camera/GPS data later.
- Forgetting that PNG screenshots compress better as JPEG when the screenshot doesn't need transparency — UI screenshots typically shrink 5-10× in JPEG.
Frequently asked questions
How does the compression find the right quality?
The encoder binary-searches JPEG quality (0-100) to find the highest setting that produces a file under the target size. It evaluates 5-10 candidates on the canvas, all without a network round-trip — the loop runs in milliseconds in your browser.
What if the source is already smaller than the cap?
The preset still re-saves it at the maximum quality the cap allows, which may be quality 95+ for small sources. The output is at most the cap size; smaller sources stay smaller.
Will compression affect the dimensions?
For most presets, no — the dimensions stay the same and only the JPEG quality drops to fit the budget. For aggressive caps (100 KB), the preset also reduces the longest edge to a maxWidth ceiling so the budget isn't spent on resolution the recipient won't display.
How does this compare to TinyPNG / TinyJPG?
TinyPNG / TinyJPG run server-side: you upload, they re-encode, you download. The presets here run entirely in your browser — no upload, no rate limit, no privacy exposure. The compression algorithms are similar (perceptual JPEG / lossless PNG) but the latency advantage is local processing.
Can I compress multiple files at once?
Yes — drop multiple files into the input area and Mochi will run the preset on each in turn, with a batch progress indicator. Outputs are saved individually.
How private is the processing?
Every utility preset runs in your browser — canvas APIs for compression, JavaScript for metadata parsing. Files containing personally identifiable information, financial details, or proprietary content stay on your device.