Upload up to 3 photos of your hardwood log. Get an instant USDA Forest Service grade — veneer, F1, F2, F3, construction, or local-use — plus a $-value estimate based on current Appalachian timber prices. Free. No signup.
Best results: one shot of the log lengthwise, one of the small end, one of the largest defect if any. Up to 3 photos.
Free • No signup • No data shared with third parties
Claude AI inspects each photo for surface defects (knots, sweep, bumps, holes, conks, cankers) and end-grain defects (ring shake, heart checks, rot, wetwood). It returns a structured defect inventory with per-defect confidence scores.
A deterministic rule engine — not a black-box ML model — applies USDA Forest Service grading rules (NE-1 for log grades, AgH 678 for defect scoring). Every grade output includes the specific rule and figure number that produced it.
Veneer, F1, F2, F3, construction, or local-use, plus a $-value range using the International ¼" log rule and current Appalachian timber prices. Share the result with buyers, foresters, or save for your records.
Surface defects with strong visual signatures (knots, sweep, large holes, conks) are detected reliably. Interior defects (heart rot, ring shake, wetwood) have no positive surface indicator per USFS NE-1 — they require physical inspection by a forester or buyer at the deck. Treat the grade as a pricing guide, not a substitute for a buyer's eyes on the log.
The deterministic rule engine implements USDA Forest Service NE-1 (Rast, Sonderman & Gammon 1973) for F1/F2/F3 factory log grades, AgH 678 (Carpenter et al. 1989) for the defect taxonomy and per-use-class scoring matrix, and the International ¼-inch log rule for board-foot scaling. Every grade output includes a reasoning trail showing which rules were applied.
Grading rules are known — codified in 50+ years of Forest Service research. We don't need ML to learn them; we need ML to see the defects so the rules can score them. The hybrid approach is explainable (every grade cites its rule), auditable (the engine's logic is open code), and impossible to hallucinate (the AI vision layer is constrained to a fixed defect taxonomy).
Yes. After grading, click "Share this grade" to get a public link like jmlogmarket.io/g/ABC12345. Send it to buyers, foresters, or post on social media. Saved grades expire after 30 days. Grades are anonymous — no email or account required.
No. A professional scaler at the deck can identify interior defects (probe for rot, sound the log for shake) that no photo-based AI can see. The AI grader is a pricing decision-support tool: it gives you a credible starting grade before the scaler arrives, helps you price a load consistently, and gives buyers context before they drive out. Use it to inform negotiations, not replace them.
Yes. No signup, no payment, no data sharing. Built and paid for by JMLogMarket as a free tool for the timber community. We make money on seller subscription plans when you list logs on the marketplace — and our hope is that a useful free tool gets you to know us before you need to list.