Non work

[limelect]

Copyright 2025 Code4Delphi Team.
Plugin designed to be used in the Delphi IDE.
Version: 2.4.0
MIT license

Gemini
Question cannot be answered
Return: {
"error": {
"code": 404,
"message": "models/gemini-1.5-flash is not found for API version v1beta, or is not supported for generateContent. Call ListModels to see the list of available models and their supported methods.",
"status": "NOT_FOUND"
}
}

chatGpt
Unable to perform processing.
REST request failed: Error getting Server Certificate

Groq
Unable to perform processing.
REST request failed: Error getting Server Certificate

Mistral
Unable to perform processing.
REST request failed: Error getting Server Certificate

Ollama It is

[egroups]

Same problem with Gemini. Cannot add new models

[limelect]

I took it out. Nice work, but useless

[Cesar4D]

Hi everyone.

Model names can be entered in the Model combobox. If the desired model isn't available for selection, simply type or paste the name and confirm, as the combobox is editable, functioning like a TEdit.

Please test it and let me know if it worked.

[limelect]

Be aware, I tried them all, and nothing works. Back to the drawing board
I personally am not going to do any research. If it does not work
flawlessly for me, it is useless

[reebopics-wq]

Hi everyone.

Model names can be entered in the Model combobox. If the desired model isn't available for selection, simply type or paste the name and confirm, as the combobox is editable, functioning like a TEdit.

Please test it and let me know if it worked.

With enter api manually the awnser of every chat is
"Question cannot be answered
Return:"

v1beta/models/gemini-3-pro-preview
v1/models/gemini-3-pro-preview
v1beta/models/gemini-2.5-flash
v1/models/gemini-2.5-flash
v1beta/models/gemini-2.5-pro
v1/models/gemini-2.5-pro
...

[reebopics-wq]

I've found a working model-name:
v1/models/gemini-2.5-flash:generateContent

[jefersoonaf]

I managed to get it working for Gemini in my case by going to the API version documentation to check if it was correct: https://ai.google.dev/gemini-api/docs/api-versions?hl=pt-br

And then, checking the current model names here: https://ai.google.dev/gemini-api/docs/models?hl=pt-br#model-versions

With this, I'm using gemini-flash-latest to always get the latest available flash model.