# TurnOver

> TurnOver is an AI-powered item valuation app. Point your phone at any secondhand item — the AI identifies it, researches what it's worth using live marketplace data, and recommends whether to sell, donate, or keep it. Built for resellers, estate liquidators, declutterers, and anyone sitting on stuff they suspect has value.

TurnOver uses a multi-model AI pipeline (Claude, Gemini, Qwen) with real-time price verification from eBay sold listings, Google Shopping, and vertical-specific sources like Scryfall (trading cards) and Discogs (vinyl records). The app covers 28 marketplace adapters including eBay, Poshmark, Mercari, Facebook Marketplace, Depop, ThredUp, Chairish, 1stDibs, Reverb, and more.

The core workflow is three steps: Snap a photo → AI analyzes and values the item → Get marketplace routing and a sell/donate/keep recommendation.

TurnOver is not a marketplace. It is the decision layer before the marketplace — helping sellers determine fair market value, choose the right platform, and understand whether an item is worth the effort to list.

## Core Pages

- [Home](https://turnover86.com/): Landing page with product overview, problem framing, feature walkthrough, and competitive positioning
- [FAQ](https://turnover86.com/faq/): Frequently asked questions about accuracy, pricing, supported items, and privacy
- [Photo Guide](https://turnover86.com/guide/): Best practices for photographing items to get the most accurate AI valuation
- [Marketplace Comparison](https://turnover86.com/marketplace-compare/): Side-by-side comparison of 28 resale marketplaces by category, fees, audience, and sell-through rates
- [Storage Cost Calculator](https://turnover86.com/storage-calculator/): Interactive tool showing the true cost of storing items vs. selling or donating them
- [Enterprise](https://turnover86.com/enterprise/): B2B tier for estate liquidators, property managers, and warehouse operations

## Articles

- [Blog Index](https://turnover86.com/articles/): All essays on recommerce, AI, and the hidden complexity of pricing secondhand goods
- [The Complexity of Pricing Anything](https://turnover86.com/articles/pricing-complexity/): Market dynamics, condition grading, regional demand, and why pricing secondhand items is genuinely hard
- [The Recommerce Shift](https://turnover86.com/articles/recommerce-shift/): Data on the growth of secondhand markets — $21T US consumption, secondhand apparel up 14%, 11.3M tons textiles landfilled annually
- [Fluent, Confident, and Often Wrong](https://turnover86.com/articles/ai-errors-part1/): Why AI errors are structural, not bugs — a plain-language explanation of hallucination, training bias, and model limitations
- [Good Enough: Why Imperfect AI Is Still Worth Using](https://turnover86.com/articles/ai-errors-part2/): Why multi-model consensus and honest uncertainty beat chasing perfect accuracy

## Legal

- [Privacy Policy](https://turnover86.com/privacy/)
- [Terms of Service](https://turnover86.com/terms/)

## Key Facts

- Categories: Electronics, furniture, clothing, collectibles, coins, trading cards, vinyl records, art, jewelry, tools, sporting goods, musical instruments, vintage items, antiques, and more
- Pricing sources: eBay sold listings (via SerpAPI), Google Shopping, Scryfall (MTG cards), Discogs (vinyl), with planned integrations for PriceCharting (games/cards) and eBay Browse API
- AI models: Claude (primary identifier), Gemini 2.0 Flash (validator), Qwen 2.5 VL 72B (scene pre-check), OpenAI (content moderation)
- Confidence scoring: Every valuation includes a confidence range and source transparency — the app tells you how certain it is and where the data came from
- Free tier: 15 AI-powered scans, then subscription pricing
- Platform: iOS (Flutter/Dart), backend on Vercel (Node.js), database on Neon PostgreSQL