Hi Retool team,
In Gemini's offering there's Gemini-2.5-flash-lite, which offer lower latency that's very useful to me, but I don't want to manually set it up everywhere (workflow, llm chat and stuff). Can we please add this?
1 Like
first go to the Retool AI Resource
then scroll to the bottom of the page
click 'Add' to add a new Custom Provider
add the google api url, basic required header values and model name
- note: you need to type the model name, then press enter or click on the dropdown to add it
and finally click the blue 'Add' button at the bottom.
- note: by selecting 'Google' for Provider Schema your api_key gets inserted by retool in the background. i think you can add
"x-google-api-key": "your_google_ai_stupid_api_key"if you wanted to use a different key for this model (not entirely sure it gets overwritten, but worth a try if it's something you need)
below you can see me using it
