To use deepseek or other models:

- Go to Chutes.ai on your browser
- Create an account
- Store your fingerprint somewhere safe (like in a txt file on google drive)
- Press the API tab or "Visit API Platform"
- Delete the default key???
- Make a new API key
- Name it whatever (doesn't matter)
- Copy it & store it somewhere safe like with your fingerprint (remember to not mix them up)
- on Janitor, go into a chat
- Click the purple button (should say "using Janitor" or similar)
- Select the Proxy tab
- There are several thigns you could put in the Model area, these ARE case-sensitive (it matters which letters are capitalized) popular choices include:
	deepseek-ai/DeepSeek-R1 - This is the old R1.
	deepseek-ai/DeepSeek-R1-0528 - This is the newest update of R1, and usually what people talk about when they say "Deepseek".
	deepseek-ai/DeepSeek-V3-0324 - The newest update of their non-reasoning models (won't make the <think> proccess, leading to quicker but less smart responses.) It's good!
	Qwen/Qwen3-235B-A22B - Qwen's biggest model, it's similar to R1 but slightly less quirky.
	tngtech/DeepSeek-R1T-Chimera - An amalgam of V3-0324 and R1 OG. It's got some very strange quirks, but can be more creative and more entertaining.
- in Other API/proxy URL, put:
	https://llm.chutes.ai/v1/chat/completions
- Put your API key in the API key field.
- For custom prompt, put whatever. I think that Janitor has a backend prompt that does *basic* instructions so it at least will know to roleplay if you don't put anything there. The prompt I use (made by myself) is linked right next to the link to this.