Frequently asked questions

It depends on the item. For common consumer electronics, kitchen appliances, and power tools, Mend is quite good — these products have well-documented failure modes and the AI has seen thousands of repair cases.

For unusual or obscure items, accuracy drops. We show a confidence indicator on every result, and when we're unsure, we say "Needs Inspection" rather than guessing. Think of Mend as a knowledgeable friend who gives you a solid starting point, not a certified technician.

This happens occasionally, especially with generic or unbranded products. Adding a photo of the model label or serial number plate dramatically improves identification. You can also type the model number in the "What's wrong?" field.

If the identification is wrong, the rest of the analysis will be off too. In those cases, try again with a clearer photo of the brand and model markings.

Mend works best with electronics, appliances, power tools, audio equipment, and mechanical devices — things with identifiable parts and repair markets. It also handles furniture, bicycles, cameras, and game consoles reasonably well.

It's less useful for clothing, soft goods, plumbing fixtures still installed in walls, or structural problems. If Mend can't identify discrete parts and price them, the analysis won't be helpful.

For DIY, we look up the specific replacement parts on eBay (both new and used sold listings) and estimate labor time based on the complexity of the repair. For professional repairs, we estimate based on typical shop rates for that category of item and the type of work involved.

These are estimates, not quotes. Actual costs vary by location, availability, and the specific condition of your item. We aim to be in the right ballpark so you can make a directionally correct decision.

Working values and parts values come from eBay sold listings — what people actually paid, not what sellers are asking. Replacement costs come from used marketplace pricing for the same or equivalent model.

We use median prices rather than averages to avoid skew from outlier sales. All prices are in USD.

No. Mend provides informal estimates for personal decision-making. The analysis is not an appraisal, not a professional assessment, and not suitable for insurance, legal, or warranty claims. If you need a documented repair estimate, take the item to a qualified technician.

Sometimes. Mend can identify many vintage items — tube amps, film cameras, mechanical typewriters — but pricing is harder because the market for vintage parts and working units is thinner and more variable. The "Needs Inspection" verdict comes up more often for vintage gear, and that's usually the right call.

For genuinely rare items, the AI might not have enough data to give useful numbers. A specialist dealer or repair shop will always know more about niche markets than a general-purpose tool.

Your photos are stored on our servers for result sharing and to improve the service. They are not sold to third parties or used for advertising. Photos are processed through OpenAI's API for the vision analysis, so they are subject to OpenAI's API data usage policy.

No account is required to use Mend. We don't collect email addresses, names, or payment information. See our privacy policy for full details.

Googling "KitchenAid mixer grinding noise" gets you to a fix eventually — after wading through forum threads, YouTube videos, affiliate blogs, and SEO filler. You'll spend 30 minutes figuring out what part you need, another 15 comparing prices, and you still won't know if the repair is worth it economically.

Mend does all of that in seconds. It identifies the part, prices it, compares repair vs. replace costs, and gives you a verdict. The goal isn't to replace repair knowledge — it's to answer the threshold question ("should I even bother?") before you invest time.

Not yet in the app, but we want to hear from you. If Mend got something wrong — misidentified an item, missed a common failure mode, or gave bad pricing — email us at hello@applesauce.chat with the result link and what was off. We use feedback to improve the system.