Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Replace GPT and Gemini with DeepSeek #1085

Open
Tash210 opened this issue Feb 7, 2025 · 12 comments
Open

Replace GPT and Gemini with DeepSeek #1085

Tash210 opened this issue Feb 7, 2025 · 12 comments
Labels
FEATURE Feature advice/request

Comments

@Tash210
Copy link

Tash210 commented Feb 7, 2025

Hi!
Some one might want to change used AI because of different reasons. So how to change AI from GPT or Gemini to Deepseek for example?

@tr1xem
Copy link

tr1xem commented Feb 7, 2025

+1 would like to see it

@rifux
Copy link

rifux commented Feb 7, 2025

Why replace? I think it's better to add a possibility to use Qwen/DeepSeek alongside with ChatGPT/Gemini

@tr1xem
Copy link

tr1xem commented Feb 7, 2025

Yes replace works too
until then zukiyajourney works fine

@Tash210
Copy link
Author

Tash210 commented Feb 7, 2025

I wanted to replace it, because GPT and Gemini are blocked in my region.

@rifux
Copy link

rifux commented Feb 8, 2025

because GPT and Gemini are blocked in my region.

But what about the other people using it?

@clsty
Copy link
Collaborator

clsty commented Feb 8, 2025

Stay chill guys. The thing is that:

  • DeepSeek API seems(?) to be compatible with OpenAI / Ollama ones, which means you're able to use it already. Though I have not tested that.
  • Functional change is not very possible happening currently since end-4 is somehow busy but PR is welcomed. And also it's not necessary to replace anything, if we're adding new function we just put them together.

@fruitsaladchan
Copy link

I wanted to replace it, because GPT and Gemini are blocked in my region.

you should be able to just add it in ~/.config/ags/services/gpt.js/

add something like this:

deepseek: {
      name: "deepseek",
      logo_name: "deepseek-symbolic",
      description: getString("Deepseek API.\nPricing: Free."),
      base_url: "https://api.deepseek.com/v1/chat/completions",
      key_get_url: "https://platform.deepseek.com/api_keys",
      key_file: "deepseek_key.txt",
      model: "deepseek-chat",
    },

Although, at least for where I'm from, they suspended their API service, so I can't really test it :p

@tr1xem
Copy link

tr1xem commented Feb 8, 2025

'deepseek': {
      name: "deepseek",
      logo_name: "deepseek-symbolic",
      description: getString("Deepseek API.\nPricing: Free."),
      base_url: "https://api.deepseek.com/v1/chat/completions",
      key_get_url: "https://platform.deepseek.com/api_keys",
      key_file: "deepseek_key.txt",
      model: "deepseek-chat",
    },

Works for me have the api key but insufficiant balance 🤣🤣🤣🤣

@clsty clsty added the FEATURE Feature advice/request label Feb 9, 2025
@Pharmaracist
Copy link
Contributor

i had added >addgpt / >rmgpt commands for my config before then followed by the model name you like from apen router it automagickly adds it to the menu without even reloading it by adding a template like you did

@IfReturn
Copy link

Maybe it is better to add a 'custom' option that allows users input url and model just like input api keys , since many ai supports openai api.

@Pharmaracist
Copy link
Contributor

when u use openrouter you just change model name and use the same other stuff

@ghl3m0n
Copy link

ghl3m0n commented Feb 18, 2025

check this out: https://ollama.com/library/deepseek-r1

then you can add it in ~/.config/ags/services/gpt.js/

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
FEATURE Feature advice/request
Projects
None yet
Development

No branches or pull requests

8 participants