Weed has a label problem: a field guide to 10 cannabis strain libraries

Read this first (disclosure). This guide is published by CannaLib.org, and CannaLib is one of the ten sites we scored. We wrote the rubric, we assigned the points, and, shocker, we did well on a rubric built around provenance and receipts. Treat the ranking as "how each site does on the things we care about," not a neutral verdict. The scores are opinion, the facts are checkable, and we show our math below. Counts for other sites are from their own pages when we checked (October 2026) and will drift. Not affiliated with any site listed.

1. The Great Strain Swindle (Or: Why Your "Sour Diesel" Might Not Be)#

You stroll into a gleaming dispensary. Apple-Store lighting. Fifty neon jars with names like Super Gary Payton, Glitter Bomb, and Granddaddy Purple. You ask for something to help you clean your apartment, and the budtender says: "This is a pure Sativa. Daytime rocket fuel."

That "Sativa" label is closer to a horoscope than a spec sheet. It might feel right, but it's a weak predictor of what the plant is genetically or chemically. DNA studies keep finding that commercial cannabis is a tangle of crosses, and that the names on the jars don't line up neatly with what's inside. Where things get messy:

• The rebrand: Slow-moving Blueberry Cookies? Relabel it with whatever went viral on TikTok.
• The clone-name roulette: Two jars of "Sour Diesel" from two shops can be quite different plants.
• The trophy tale: Seed banks love to say "Cannabis Cup winner." CannaLib checked 4,826 of those claims against 21,141 competition results (264 competitions, 61 countries, 1988 to 2026): 1,599 confirmed, 1,070 not in the results, 167 with the wrong year, 41 where a parent was the real winner, and the rest too vague to pin down.

So a wave of strain libraries has appeared to sort out fact from vibe. Some are family trees, some are DNA labs, some are Yelp with terpenes. Here's the field guide.

2. Cheat Sheet: What Actually Matters#

1. Cannabinoid types (chemotypes). Type I: THC-dominant, little CBD (most recreational flower). Type II: balanced THC and CBD. Type III: CBD-dominant.
2. Terpenes. The aroma molecules (pine, pepper, gas, tropical fruit). Researchers still argue about how much they steer effects, so treat them as a clue, not a guarantee. A decent rule of thumb is that flower with higher total terpenes tends to smell stronger.
3. Receipts. Where did it come from, who said so, and did anyone actually check?

3. The Scorecard#

We scored ten libraries from 0 to 100 using six weighted categories: Provenance and citations (20%), Lineage depth (20%), Chemistry and lab data (20%), Catalog scale and cleanliness (15%), Grower field data (15%), Usability and tech (10%). Re-weight them yourself if you disagree. (Scores are CannaLib's opinion.)

4. The Contenders#

#1. CannaLib.org: Receipts-First (and Yes, It's Us)

• Live numbers (October 2026): 143,611 named strains · 7,358,743 evidence-backed data points · 1,276,466 lab tests · 118,039 lineage links
• Links: Home | Explorer | Lineage map

CannaLib's motto is "Every strain, every claim, with its receipt." Most strain sites quietly pick the best-sounding parentage and average the THC numbers into one friendly figure. CannaLib has four Honesty Rules:

1. Never invent. No data? The field stays blank.
2. Never blend. Three seed banks, three THC numbers? You see all three. (Sources disagree on THC for 37,014 strains.)
3. Never merge. Strains don't quietly inherit a parent's stats.
4. Never shred. Ranges stay ranges.

• Cup claims, checked: see above. 1,599 of 4,826 confirmed.
• Lineage map: 57,015 strains in generation rings from landraces outward.
• Catalog scrub: 120,822 product and label names ("Strawberry Cough 0.5g Cartridge") kept out of the strain count.
• The honest caveat: it's a research library. No shopping cart, and a lot of what's in it is claims, not proof. Only 1,017 strains have a real DNA sample behind them.

#2. SeedFinder.eu: The Ancient Family Scrolls

• Numbers (site, Oct 2026): about 41,700 strains · 2,142 breeders · 106 "pure local breeds" (landraces) · ThreadFinder across 35 forums

The grand old archive of seed-bank genetics: parents, grandparents, backcrosses, and reversals, going back to landraces. Want to know what happened when a famous Skunk met a backpack Afghani? Probably here. The catch: a dated interface, and it largely reports what breeders say, which is the nature of the beast.

#3. Kannapedia: The CSI Lab

• Numbers (site, Oct 2026): about 2,400 registered strains (2,411 when we looked)

Run by Medicinal Genomics, Kannapedia reads actual DNA. If a famous name is stuck on a cheap clone, genetics can often tell. It's a gated catalog (somebody has to submit a sample) and the reports reward science fans, but it's the closest thing to ground truth here.

#4. GrowDiaries: Instagram for Soil Nerds

• Numbers (site, Oct 2026): about 8,180 strains listed

Seed catalogs say "600 g/m², flowers in 7 weeks, smells like unicorn candy." GrowDiaries is where growers post the unfiltered week-by-week photos. Great reality check before you pop seeds; not much for the dispensary shopper. Fun detail: its Strain of the Year votes are even in CannaLib's competition data.

#5. Leafly: The Slick Mall of Weed

Launched in 2010. In October 2019 it introduced its Cannabis Guide with color-coded terpene "flowers", moving away from the old indica/sativa/hybrid tiles toward chemistry. Easy to use and tied to local ordering; the trade-off is commercial weight (sponsored placements) and strain profiles that show an average rather than your actual batch.

#6. Hytiva: The Glossy Centerfold

Studio-quality macro photography plus lab breakdowns. A beautiful, smaller catalog. Check its site for current retail coverage; availability has been limited to some markets.

#7. Jane Technologies: The Barcode Brain

Plugs into dispensary menus and inventory so what you see is what's on the shelf, with batch lab data where the store provides it. A shopping engine, not an encyclopedia.

#8. Phylos Galaxy: The Sci-Fi Star Map

Phylos Bioscience launched the Galaxy in April 2016: a 3D map where each point is a sequenced sample and related plants cluster together. It showed the public that names and genetics don't always match. Then, in 2019, Phylos's plan for a breeding program caused a backlash among growers and led to the Open Cannabis Project dissolving.

#9. AllBud: The Medicine Cabinet

Symptom-first browsing for patients. Useful for "what do people try for X?", dated in layout, light on lineage.

#10. Weedmaps: The Deal Hustler

A giant marketplace first, a strain library second. Great for "who has it near me," thin on botany.

5. Grudge Matches#

• "Where did this strain come from?" SeedFinder for breeder-lore, Kannapedia for DNA, CannaLib for the cross-source comparison with receipts. Use two of the three.
• "I just bought seeds. What now?" GrowDiaries.
• "Great weed tonight." Leafly to browse, Jane to find what's actually in stock.

6. Superlatives (Not Medals)#

• Best breeder family trees: SeedFinder.eu
• Best DNA ground truth: Kannapedia
• Best grower reality check: GrowDiaries
• Best consumer browsing: Leafly
• Best live-inventory accuracy: Jane
• Best photography: Hytiva
• Best claim-checking and cross-source comparison: CannaLib.org (our category, obviously)

The Smoker's Rule of Thumb#

1. Smell it first. Ignore the THC headline. Terpene numbers and your nose tell you more.
2. Check the receipts. If a brand says "10 Cannabis Cups," look it up before you pay for it. CannaLib's award audit is one place to start.
3. Growing? Read a few journals on GrowDiaries first.

The plants haven't changed. The tools for checking the stories about them finally caught up.