Additional providers
johnd0e opened this issue · comments
It would be nice to have also:
- gemini-pro: https://blog.google/technology/ai/google-gemini-ai/
Official SDK: https://github.com/google/generative-ai-python - copilot Bing: https://copilot.microsoft.com/
Libs: - copilot Github: https://github.com/features/copilot
Sample code: https://github.com/CaoYunzhou/cocopilot-gpt - Lib: https://github.com/xtekky/gpt4free (Many subproviders)
- https://wrtn.ai/ (GPT-4, Claude 2.1 etc)
Sample code: https://github.com/cannonLCK/wrtn_python - Claude https://claude.ai/chats
Libs: - https://huggingface.co/chat/ (Available Models: Mixtral-8x7B-Instruct-v0.1, Llama-2-70b-chat-hf, Nous-Hermes-2-Mixtral-8x7B-DPO, CodeLlama-34b-Instruct-hf, Mistral-7B-Instruct-v0.2, openchat-3.5-0106)
Lib: https://github.com/Soulter/hugging-chat-api - NotionAI: https://www.notion.so/product/ai
Lib: https://github.com/vaayne/notionai-py
Could you provide links to these models.
I would like to highlight the following.
Google's Bard powered by Gemini is already implemented though I'm not sure if it's pro.
OpenAI's GPT4 is already implemented too. Thanks to WebChatGPT. It's accessed slightly different $ pytgpt webchatgpt
Could you provide links to these models.
Done
Google's Bard powered by Gemini is already implemented though I'm not sure if it's pro
Google Bard is still powered by PaLM worldwide, only in US it is by Gemini.
And current Bard support implementation is far less reliable than straightforward gemini api usage via Google SDK.
OpenAI's GPT4 is already implemented too. Thanks to WebChatGPT
Well, GPT-4 via web chat is available only with Plus subscription.
But I actually mean https://github.com/xtekky/gpt4free, it is lib with many sub-providers.
but one of it's major con is the dependency on Google Chrome
Most providers does not require it at all.
Some can get cookies out of it (but it is also possible to provide cookies via arguments).
And that's why we shall have to recreate them to achieve the said functionality.
Line 10 in c90c6dc
You probably should not fix any particular version, as g4f is very unstable project: frequent updates, and frequent providers changes. Yesterday some providers were ok, but today they may just gone (possibly forever).
Also consider Status column of this table: https://github.com/xtekky/gpt4free/tree/main?tab=readme-ov-file#-providers-and-models: there are plenty providers, but only 12 marked as Active at the moment.
Just updated the list (in the first post)
Bumping this issue because it has been open for 14 days with no activity. Closing automatically in 14 days unless it becomes active again.
I think we should add Gemini-pro api support since it's free upto 60 requests per minute. You need an api-key which can be generated by signing into Google AI Studio. Also check this out https://github.com/llegomark/gemini-pro-chat.
Okay I will look into it.
I think we should add Gemini-pro api support since it's free
Bumping this issue because it has been open for 14 days with no activity. Closing automatically in 14 days unless it becomes active again.
Hello, I am not sure if it is the right place to ask but is there a way to use the g4f models other than the 13 first ones in the readme in python ? I can't find a way to use them along with pytgpt.
Bumping this issue because it has been open for 14 days with no activity. Closing automatically in 14 days unless it becomes active again.
Keep it aliv
Hello, I am not sure if it is the right place to ask but is there a way to use the g4f models other than the 13 first ones in the readme in python ? I can't find a way to use them along with pytgpt.
Try out this
import pytgpt.gpt4free as gpt4free
bot = gpt4free.GPT4FREE(provider="Koala")
print(bot.chat("<Your-prompt>"))Bumping this issue because it has been open for 14 days with no activity. Closing automatically in 14 days unless it becomes active again.
Closing due to inactivity.