DeepSeek VisionPlus
qq247505/DeepSeek-VisionPlus · model-adapters
DeepSeek Harness Visual Plugin - Text goes DeepSeek Official API, photo automatic route to the free visual model pool.
★ 2GitHub stars
0Forks
2026-08-25Last updated
TypeScriptLanguage
MITLicense
Key features
- Automatic visual routeautomatically hand over to the visual pool when a picture (selection / pasting / model autonomy read image) appears in the message, DeepSeek is still requested without a chart and Off/ High/ Max is retained as it stands
- Set up storageConfigure the namespace of the plugin, read and write via the plugin's own mounted interface (/api/visionPlus.settings) - without llm-pi-ai being borrowed, internal lines will not appear in the model selector
- Test failedKey invalid (401: check if model key is completed correctly
- Set up cardSetup Bridge Visual DeepSeek VisionPlus cards in the stand-alone field - Add visual models, configure keys, edit model directories (ID/ name/ context window/ maximum output, K/M units) to save real time effective
- Enter box picture buttonShow up after selecting visual variants, selected / multiple selection / drag and drop
Install command
dsh plugin --profile web add github:qq247505/DeepSeek-VisionPlus