Model zoo
Every model AddaxAI ships, in one searchable table. Filter by type, filter by region, or search by name, developer, or region. This table is generated from the app's real model catalogue (models.json), so it always matches what the app installs.
39 models
| Model | Type | Developer | Region | Summary | Min app | Links |
|---|---|---|---|---|---|---|
MegaDetector v5a MD5A-0-0 | Detection | Dan Morris | — | Industry standard • Battle tested | 7.0.1 | InfoLicenseCite |
MegaDetector v5b MD5B-0-0 | Detection | Dan Morris | — | Alternative to 5a • Dataset-dependent performance | 7.0.1 | InfoLicenseCite |
MegaDetector v1000 Redwood MD1000-REDWOOD-0-0 | Detection | Dan Morris | — | Promising new standard • Not battle tested yet | 7.0.1 | InfoLicenseCite |
MegaDetector v1000 Spruce MD1000-SPRUCE-0-0 | Detection | Dan Morris | — | Low accuracy (-14%) • 13x faster than 5a | 7.0.1 | InfoLicenseCite |
MegaDetector v1000 Cedar MD1000-CEDAR-0-0 | Detection | Dan Morris | — | High accuracy (-1%) • 2x faster than 5a | 7.0.1 | InfoLicenseCite |
MegaDetector v1000 Larch MD1000-LARCH-0-0 | Detection | Dan Morris | — | Good accuracy (-3%) • 2.4x faster than 5a | 7.0.1 | InfoLicenseCite |
MegaDetector v1000 Sorrel MD1000-SORREL-0-0 | Detection | Dan Morris | — | Moderate accuracy (-3%) • 7x faster than 5a | 7.0.1 | InfoLicenseCite |
🌏SpeciesNet 4.0.2a SPECIESNET-v4-0-2-A | Classification | global | Global • 2,000+ classes • Google Research | 7.0.1 | InfoLicenseCite | |
🇪🇺Deepfaune v1.3 EUR-DF-v1-3 | Classification | The DeepFaune initiative | europe | Europe • 34 classes | 7.0.1 | InfoLicenseCite |
🌵Sub-Saharan Drylands SAH-DRY-ADS-v1 | Classification | Addax Data Science | africa | East & Southern Africa savannas • 328 classes • Trained on 2.8M images | 7.0.1 | InfoLicenseCite |
🇳🇵Terai region (Nepal) TERRAI-NEP-v1 | Classification | Alexander Merdian-Tarko | asia | 10 classes • Tiger conservation | 7.0.1 | InfoLicense |
🇦🇺Tasmania TAS-BB-v1 | Classification | Barry Brook | oceania | 96 classes • Trained on 2.5M images | 7.0.1 | InfoLicenseCite |
🇳🇦Namibia (Skeleton Coast) NAM-ADS-v1 | Classification | Addax Data Science | africa | 30 classes • Trained on 850k images | 7.0.1 | InfoLicenseCite |
🇳🇿New Zealand (invasives) NZI-ADS-v1 | Classification | Addax Data Science | oceania | 17 classes • Trained on 2M images | 7.0.1 | InfoLicense |
🇵🇪Peruvian Amazon (rainforest) PAM-SDZWA-v1 | Classification | San Diego Zoo Wildlife Alliance | americas | 53 classes • San Diego Zoo Wildlife Alliance | 7.0.1 | InfoLicense |
🇹🇲Turkmenistan TKM-ADS-v1 | Classification | Addax Data Science | asia | Southern region • 14 classes • Trained on 1M images | 7.0.1 | InfoLicenseCite |
🇰🇬Kyrgyzstan KIR-HEX-v1 | Classification | Hex Data | asia | Manas v1 • OSI-Panthera • Trained on 42k images | 7.0.1 | InfoLicense |
🇺🇸Southwest USA SWUSA-SDZWA-v3 | Classification | San Diego Zoo Wildlife Alliance | americas | Version 3 • 27 species • Trained on 92k images | 7.0.1 | InfoLicense |
🇯🇵Gifu region (Japan) GIF-JAP-v0-2 | Classification | Gifu University | asia | 13 classes • prototype on limited, imbalanced data • very biased to deer | 7.0.1 | InfoLicenseCite |
🇺🇸Hawaiʻi HWI-ADS-v1 | Classification | Addax Data Science | americas | AI Puaʻa v1.0 • 15 classes • USDA Forest Service & TNC | 7.0.1 | InfoLicense |
🇦🇺Victoria (Australia) VIC-ADS-v1 | Classification | Addax Data Science for Parks Victoria | oceania | 212 classes • Trained on 5M images | 7.0.1 | InfoLicense |
🇺🇸US Southwest (borderlands) SBUSA-ADS-v1 | Classification | Addax Data Science | americas | 68 classes • 3M training images | 7.0.1 | InfoLicense |
🇺🇸Midwest US (drift fences) AHDRIFT-v1 | Classification | The Ohio State University, Columbus Zoo and Aquarium, Addax Data Science | americas | Built for top-down close-up cameras | 7.0.1 | InfoLicense |
🇮🇳Central India IND-ADS-v1 | Classification | Addax Data Science | asia | Fine-tuned SpeciesNet • 40 classes • Trained on 1.1M images | 7.0.1 | InfoLicense |
🇦🇺Top End savanna (Australia) ANT-ADS-v1 | Classification | Addax Data Science | oceania | Fine-tuned SpeciesNet • 140 classes • Trained on 890k images | 7.0.1 | InfoLicense |
🇪🇺Deepfaune v1.4 EUR-DF-v1-4 | Classification | The DeepFaune initiative | europe | Newest Deepfaune • 38 classes • 50+ European partners | 7.0.1 | InfoLicenseCite |
🇪🇺Deepfaune v1.2 EUR-DF-v1-2 | Classification | The DeepFaune initiative | europe | Older Deepfaune release • 30 classes • for reproducing past runs | 7.0.1 | InfoLicenseCite |
🇪🇺Deepfaune v1.1 EUR-DF-v1-1 | Classification | The DeepFaune initiative | europe | Older Deepfaune release • 26 classes • for reproducing past runs | 7.0.1 | InfoLicenseCite |
🇵🇪Peruvian Andes PAN-SDZWA-v1 | Classification | San Diego Zoo Wildlife Alliance | americas | Andean highlands • 53 classes | 7.0.1 | InfoLicense |
🇦🇺Queensland Wet Tropics QLD-WOB-v1 | Classification | Prakash Palanivelu Rajmohan and Renuka Sharma | oceania | Wet Tropics rainforest • 15 classes | 7.0.1 | InfoLicense |
🇦🇺Australia (AWC135) AWC135-AWC-v1 | Classification | Australian Wildlife Conservancy | oceania | Continent-wide • 135 classes | 7.0.1 | InfoLicense |
🇳🇿New Zealand (species) NZS-WEK-v3-03 | Classification | wekaResearch | oceania | Native + introduced • 81 classes | 7.0.1 | InfoLicense |
🌴Neotropics (TropiCam-AI) NEO-MNCN-v1-0 | Classification | Andrea Zampetti, National Museum of Natural Sciences (MNCN-CSIC), Madrid, Spain | americas | Central + South America • 84 classes | 7.0.1 | InfoLicense |
🌳African tropical forests v1 AFR-DFV-v1 | Classification | Hugo Magaldi - One Forest Vision initiative | africa | Congo Basin rainforest • 34 classes | 7.0.1 | InfoLicense |
🌳African tropical forests v2 AFR-DFV-v2 | Classification | Hugo Magaldi - One Forest Vision initiative | africa | Congo Basin rainforest • 61 classes • DINOv3 | 7.0.1 | InfoLicense |
🇨🇴Colombian Amazon CAM-AI4G-v1 | Classification | AI For Good Lab, Microsoft | americas | Amazon rainforest • 36 genera • PytorchWildlife | 7.0.1 | InfoLicenseCite |
DINOv2 ViT-S/14 DINOV2-VITS14 | Embedding | Meta AI (FAIR) | — | Fastest • 384-dim • Recommended for most projects | 7.0.1 | InfoLicense |
DINOv2 ViT-B/14 DINOV2-VITB14 | Embedding | Meta AI (FAIR) | — | Balanced • 768-dim • ~2.5x slower, finer detail | 7.0.1 | InfoLicense |
DINOv2 ViT-L/14 DINOV2-VITL14 | Embedding | Meta AI (FAIR) | — | Best quality • 1024-dim • ~7x slower, GPU recommended | 7.0.1 | InfoLicense |
Model types
- Detection: finds animals, people, and vehicles in an image and draws boxes around them. MegaDetector is the default. Detection runs first, on every file.
- Classification: names the species inside each animal box. SpeciesNet is the global default; regional classifiers cover specific ecosystems.
- Embedding: turns each detection crop into a feature vector for similarity search and clustering, so you can find visually similar detections fast.
The "Min app" column is the minimum AddaxAI version that can run the model. If a model needs a newer version than you have installed, update the app.