Next Todos and Maybes

[hlohaus]

Please give Feedback :)

• New Google Bard provider without webdriver, but with image gen and upload ✅
• Update the provider list with the tool ✅
• Add a example for the image generator's in the readme ✅
• Remove required auth check for all providers or only for providers with webdriver login ✅
• Blog Post about g4f on Smartphone✅
• New Provider: https://chat.deepseek.com/coder
• Add Python API for creating images ✅

Maybes:

• CLI interface for ChatCompletion
• Python API like from OpenAi working ✅
• Better example for ChatCompletion / history support
• new API for Conversations / create conversation / too long message in OpenaiChat issue
• or/and Disable History Button in GUI ✅
• Auth for providers with GPT 4 and Auth: You ✅, Phind..
• Add more image gen providers
• Check and improve bot detection bypass for OpenaiChat ✅ and Bing
• Add provider: https://flowgpt.com/✅
• Remove allowed models check in Completion API?
• Bard Speech Api, OpenAi Speech Api
• Improve Error Messages: -- MissingGoogleChromeError -- MultipleWebDriverSessionError -- TooLongMessageOpenAiError

Also should we add account generation? It is too silly? today, all providers with GPT 4 requiring a account.

• Added You provider with account generation ✅

[OmiiiDev]

If you need I have plenty image Gen providers that I reverse engineered

[hlohaus]

Hello, yes, please. I have added image gen to the readme. With Gemini: we have 3 image providers.

[OmiiiDev]

Alright but it's a lot of lines of code and providers so how can I send it to you? @hlohaus

[hlohaus]

Can't you create a repo? Or you create a PR with your draft?

[OmiiiDev]

Not really, because I made it a flask app with ready api and everything so I might not be able to share it easily

[hlohaus]

Okay, write me a Email to: ..

[OmiiiDev]

Alright will send the email asap

[OmiiiDev]

I emailed you with the code snippets

[hlohaus]

I have your email. I could copy your generate_image to g4f.Provider.NewImgGen and we a have new Provider. Good idea?

[OmiiiDev]

Yes sounds good

[OmiiiDev]

I reverse engineered a tts model, so if you need it I can send it aswell

[thangckt]

can you include references for Phind response?

[hlohaus]

The provider g4f.Provider.Phind are working, but only they Phind model are supported.

[github-actions[bot]]

Bumping this issue because it has been open for 7 days with no activity. Closing automatically in 7 days unless it becomes active again.

[freedomforwhat]

Hello! Is it possible to add a check for a new version of the repository when starting the GUI and when starting any chat in the GUI?

[skoenig]

Would be really nice to include a small CLI tool to enter (or pipe) prompts to a model and receive the answer via the command line. Some basic options besides provider and language model could be: context ID (to pick up the conversation where you left), initial prompt, load and modify chat history.

[plia7]

• CLI interface for ChatCompletion

Yes, @hlohaus is it possible to add CLI interface for ChatCompletion please? Thanks, it's a great project.

[TungusSs]

Is it possible? #1766

[thisisjerk]

Is it possible to include a conversation rename feature in the left panel?

[hlohaus]

Yeah, that's possible. I'm doing that.

[Hydra566]

I can suggest pollinations.ai as an image provider.

Images are generated quickly and the quality is good. Powered by SDXL.

Queries are also easy to create:

https://image.pollinations.ai/prompt/dog,%20high%20quality,%20high%20detail,%20hyperrealism,%208k,%2016k?width=1024&height=1024&seed=43&nofeed=true&nologo=true

To get a new image, you need to replace seed=43 with any random number, for example seed=45645673457.

[github-actions[bot]]

Bumping this issue because it has been open for 7 days with no activity. Closing automatically in 7 days unless it becomes active again.

[github-actions[bot]]

Closing due to inactivity.