Builkding a n8n VisionAI connector - Detect object and more

Summary

We continued building the Reka AI connector for N8N by adding brand-new vision operations: prompt about an image, prompt about a short video, and detect objects in an image, powered by the Reka Edge model. Plenty of debugging along the way as we figured out how to properly structure the API calls and route between the different media types inside the connector!

📺 - Twitch archive - stream no. 410

Replay

Project

All the code for this project is available on GitHub: n8n-nodes-reka - https://github.com/FBoucher/n8n-nodes-reka

Game Results

• @fredda_the_cat: 72.94
• @phrakberg: 67.53
• @surlydev: 12.31
• @fboucheros: 0
• @jtsom: 0
• @groversaurus: 0

Statistics

• 🏆Best score: @fredda_the_cat with 72.94
• 😭Biggest loser: @groversaurus with 11 drops and no high score
• 🍀Luckiest: @fredda_the_cat with best score 72.94 and only 11 drops
• 🎖️Super participant: @fredda_the_cat with 11 drops