How to set and use properly configuration file?

[aborruso]

Hi @ErikBjare,
once again sorry for this basic question, but this is a great tool, I would like to use it to the best of my ability, and I still don't really understand some basic things.

Below an example configuration file.

Some questions:

• If I set the API key for multiple models, which one is used as the default model? Do I have to comment them all out and leave only one free?
• for one API key I would use one model, for another key another model, but if I set two "MODEL" rightly gptme goes in error. The correct way to handle this is to choose the model at launch with --model and based on the model the right KEY API will be chosen?
• and for OPENAI_BASE_URL, I only have to set it if I use ollama, and in all other cases will the URL be changed according to the model chosen?

Thank you

[prompt]
about_user = "I am a curious human programmer."
response_preference = "Basic concepts don't need to be explained."

[prompt.project]
activitywatch = "ActivityWatch is a free and open-source automated time-tracker that helps you track how you spend your time on your devices."
gptme = "gptme is a CLI to interact with large language models in a Chat-style interface, enabling the assistant to execute commands and code on the local machine, letting them assist in all kinds of development and terminal-based work."

[env]
ANTHROPIC_API_KEY = "sk-ant-api03-xxxxxxxx5gmAAA"
GROQ_API_KEY = "gsk_FxxxxxwxxxxxxkX"
OPENAI_API_KEY = "gsk_CIxxxxxRHuz"
ANTHROPIC_API_KEY = ""
OPENROUTER_API_KEY = ""
XAI_API_KEY = ""
GEMINI_API_KEY = ""
DEEPSEEK_API_KEY = ""

MODEL="groq/llama-3.1-8b-instant"
MODEL = "local/llama3.1:latest"

#OPENAI_BASE_URL = "https://api.groq.com/openai/v1"

# Uncomment to use with Ollama
#OPENAI_BASE_URL = "http://localhost:11434/v1"

[ErikBjare]

If I set the API key for multiple models, which one is used as the default model? Do I have to comment them all out and leave only one free?

If no model is set, it will try to detect a provider with keys configured:

https://github.com/ErikBjare/gptme/blob/54253a5db4ef37175402a650ba759ed842831619/gptme/llm/__init__.py#L273-L293

You can set multiple keys and use MODEL or -m/--model to chose a specific one.

for one API key I would use one model, for another key another model, but if I set two "MODEL" rightly gptme goes in error. The correct way to handle this is to choose the model at launch with --model and based on the model the right KEY API will be chosen?

Yes. The model prefix like anthropic/... decides the key used.

and for OPENAI_BASE_URL, I only have to set it if I use ollama, and in all other cases will the URL be changed according to the model chosen?

Yes. Only models with the local/... provider prefix respect the OPENAI_BASE_URL (which in hindsight may have been a mistake, since OPENAI_BASE_URL may not be set to something local, but it works).

https://github.com/ErikBjare/gptme/blob/54253a5db4ef37175402a650ba759ed842831619/gptme/llm/llm_openai.py#L66-L73

[aborruso]

You can set multiple keys and use MODEL or -m/--model to chose a specific one.

Hi @ErikBjare if I set multiple models

MODEL="groq/llama-3.1-8b-instant"
MODEL = "local/llama3.1:latest"

I have tomlkit.exceptions.KeyAlreadyPresent: Key "MODEL" already exists.

If I set multiple keys and models, what's the default one?

Thank you

[ErikBjare]

You cannot set multiple MODELs, you can set multiple keys.

So to rephrase what I said before:

You can set multiple keys, and set MODEL once to set a new default, or -m/--model to chose a specific one for that session.

[aborruso]

I'm stupid, sorry. You have written well and clearly.

Thank you again