base_url #1

New Issue

GiteaMirror · 2025-11-06T14:30:59-06:00

GiteaMirror commented

2025-11-06 14:30:59 -06:00

Originally created by @comeback01 on GitHub (May 1, 2024).

Originally assigned to: @Shubhamsaboo on GitHub.

Hello,

I would like to know how to change the base URL. The idea is to be able to use OpenRouter, which is compatible with OpenAI, and they offer free models.

Thank you very much!

Originally created by @comeback01 on GitHub (May 1, 2024). Originally assigned to: @Shubhamsaboo on GitHub. Hello, I would like to know how to change the base URL. The idea is to be able to use OpenRouter, which is compatible with OpenAI, and they offer free models. Thank you very much!

GiteaMirror closed this issue

2025-11-06 14:30:59 -06:00

GiteaMirror commented

2025-11-06 14:31:00 -06:00

@Shubhamsaboo commented on GitHub (May 4, 2024):

Can you please elaborate on this?

Which example exactly you are referring to?

@Shubhamsaboo commented on GitHub (May 4, 2024): Can you please elaborate on this? Which example exactly you are referring to?

GiteaMirror commented

2025-11-06 14:31:00 -06:00

@comeback01 commented on GitHub (May 4, 2024):

There are suppliers that offer APIs that are 100% compatible with OpenAI, such as:

I particularly like https://openrouter.ai because it provides models like Claude 4, GPT-4, Meta: Llama 3 70B Instruct, etc.

This provider also offers completely free models, such as Mistral 7B Instruct.

To use these models, there is no need to modify the code. However, you need to specify:

an API key (api_key)
the chosen model (model)

As in your program, you must also provide a base URL (baseURL), which could be:

I hope this makes things clearer.

Thank you for your help.

@comeback01 commented on GitHub (May 4, 2024): There are suppliers that offer APIs that are 100% compatible with OpenAI, such as: - https://www.together.ai/ - https://fireworks.ai/ - https://openrouter.ai I particularly like https://openrouter.ai because it provides models like Claude 4, GPT-4, Meta: Llama 3 70B Instruct, etc. This provider also offers completely free models, such as Mistral 7B Instruct. To use these models, there is no need to modify the code. However, you need to specify: - an API key (api_key) - the chosen model (model) As in your program, you must also provide a base URL (baseURL), which could be: - https://openrouter.ai/api/v1 - https://api.together.xyz/v1 I hope this makes things clearer. Thank you for your help.

GiteaMirror commented

2025-11-06 14:31:00 -06:00

@Shubhamsaboo commented on GitHub (May 5, 2024):

Gotcha.

Embedchain supports the use of following LLM and support the providers listed here: https://docs.embedchain.ai/components/llms

All you need to do is to change the config of your embedchain app. You can find the instructions on how to use them on the above link.

I hope this solves your problem.

@Shubhamsaboo commented on GitHub (May 5, 2024): Gotcha. Embedchain supports the use of following LLM and support the providers listed here: https://docs.embedchain.ai/components/llms All you need to do is to change the config of your embedchain app. You can find the instructions on how to use them on the above link. I hope this solves your problem.

GiteaMirror commented

2025-11-06 14:31:00 -06:00

@Shubhamsaboo commented on GitHub (May 17, 2024):

Closing this as there is no response for a while.

@Shubhamsaboo commented on GitHub (May 17, 2024): Closing this as there is no response for a while.

GiteaMirror referenced this issue

2025-11-06 14:59:29 -06:00

[PR #208] [MERGED] add EvoAgentX to multi agent app #243

Sign in to join this conversation.