8 actions in 8 pieces of software · ranked by words, meaning and measured trust · 593 ms
modelrecord20,137 downloads/month
- zero-shot-object-detection
Task: zero-shot-object-detection
- zero-shot-object-detection
modelrecord29,976 downloads/month
- image-classification
Task: image-classification
- image-classification
modelrecord1,161,513 downloads/month
- image-feature-extraction
Task: image-feature-extraction
- image-feature-extraction
modelrecord1,473,280 downloads/month
- zero-shot-image-classification
Task: zero-shot-image-classification
- zero-shot-image-classification
modelrecord
- adirik/owlvit-base-patch32
Zero-shot / open vocabulary object detection
- adirik/owlvit-base-patch32
modelrecord180,867 downloads/month
- video-classification
Task: video-classification
- video-classification
modelrecord
- lucataco/clip-vit-base-patch32
openai/clip-vit-large-patch32
- lucataco/clip-vit-base-patch32
- 8obsidian-mcpfound in code, not yet checkedno passport yet
MCP serverykoellmann/obsidian-mcp
- patch_base_tool
Atomically update an existing .base file without rewriting it wholesale. update_formulas/update_properties are merged by key. set_filters replaces the whole filters block. update_views: each dict must include 'name'. data: {views} (count).
- patch_base_tool
Labels are the instruments' verdict and are never changed by ranking. Agents and robots get this same answer as JSON at /api/v1/find?q=…&answer=1 and through the MCP server.