AvaniJB - Total Latte (and Gemini) Love

Now also for gemini-pro-2.0 and GPT 4.5

Avani

JB made to be easy to edit/customise and extremely lightweight at less than 200 tokens for all essentials.

Everything you need to know is in the FAQ and JB. If something doesn't work, @ on discord, check the thread on Scylla. Burner is littledoodler@proton.me. I'm on /aicg/ sometimes. All art is by myself, not genned.

Credits

all my testers and helpers!!! (permanent credits: @kira####, @pur####, @holyw####, @sopade####, @hal####, @guil####)
karukaru, knottydog, mongo, and whoever else for the new del prompt for 2.4. I didn't come up with the idea, i merely adapted it (because frankly for this latte version it works the best)
koalache & karukaru for exchanging ideas and helping me change dumb stuff. Also being fellow jbmakies
@rarest#### for being an absolute regex entity, and giving me a few ideas
@Sir#### for access to his proxy, as well as donating a regex and his info board :lattelove:
Mr. Regex :salute:
@minsk#### for his amazing JED jb, and exchanging ideas with
@Smiley####, for inspiring previous presets!
the lattelover cult on scylla :lattelove:
and someone who doesnt want to be named but got me into making jbs now, i guess. theyre also to be credited for at least 1/3rd of the logit bias.

DOWNLOAD HERE

2.5 - Universal Edition | Updated 06.02.25

My stuff seems to work on other things, including Nonnet/Sorbetto/Sonnet 3.6 and Deepseek. So if you need something less specialised, here you go.

catbox fileditch

2.5.1 - Filter Spray Edition (also 4.5 friendly) | Updated 10.03.25

Written for both chatgpt-4o-latest and any 4o versions. I use chatgpt-4o-latest at the moment, the snapshot kinda sucks ass. Yes, it also works for other GPT-4o versions, you may need to tweak the logit, though. The 4.5 version is seperate due to needing seperate settings. The seperate settings will get ported to Latte eventually, but if you want to use it for Latte, just turn on squash system messages.

WD40

catbox, latte fileditch, latte

catbox, 4.5 fileditch, 4.5

Logit Bias (v14.1) — last updated 07.02.25 — NOT optional, 100% required. Works for 4.5.

catbox fileditch

Logit Bias (clean), works for 4.5 — last updated 27.02.25

If you want a version of my Logit Bias that is FAR FAR FAR FAR easier to edit and get an overview of, here you go. It's part of an ongoing rework to clean it up and make it usable beyond personal purposes. You can delete the headers if you don't need them, they take up logit bias space, but there's no better way of formatting things. If you get too many <, turn down <. It does however help with filter evasion. Turn it up in decimals (eg. 4.5, 4.6, etc.) if you get filtered. Note: the token I is set at -100 for filtering purposes. Turn it down if you don't need it.

catbox fileditch

Here's also a version of this that's currently being adjusted to personal preference, aka with lots of words that I hate, that's also easier to edit.

catbox fileditch

yes you can use this with whatever other preset you want. not stopping you.

Logit Pro Tips

Change the value of . to have shorter/longer sentences. Shorter = Higher positive value (e.g. 1), Longer = Negative value (e.g. -1.)
Change the frequency of paragraphs by changing the value of [364]. However, this can only do so much, so I made a version for people who like their paragraphs SUPER long: Long Paragraph Logit
The issue with paragraphs is that GPT will obviously just replace it with stuff like double space and then paragraph break, which is also one token. I'm sure you can eventually get consistently one long paragraph if you simply ban every other form of paragraph break lol
To check the number of a token, type it into the Token Counter in ST, which is on the magic wand next to the bottom left menu. It also helps to turn on request token probabilities from settings, and then check them in token probabilities from the bottom left menu. This is helpful to check tokens that aren't immediately visible, like paragraph breaks, or the percentage chance for individuals tokens.
Change the value of , to diversify your writing more. If you notice more GPTisms than usual or there's way too many run on sentences, logiting out , is a pretty good way to outright kill some of them.
Generally, if there is a phrase that you hate, it's not enough to just delete the offending word. Example: Paragraphs starting with 'The silence'. If you logit out silence, it's just going to replace silence with something similar, like quiet, or the famous mispellings. If you logit out The and This and That (and similar sentence starts), you diversify the token choice and give it the opportunity to start with a token that doesn't lead into describing silence. Basically, you want to murder the structure, not the individual words (usually, although sometimes doing the latter DOES help!).
Logit helps break the filter AND makes writing more diverse. I suggest experimenting a little with the things you can do. I usually use Logit Bias as a way to diversify my writing mid-way through an RP, which helps keep it fresh.

Regexes

Your required regexes depend on what you want to use. The non-End regexes are optional QoL. Credit for new delete error and delete last paragraph regex goes to Mr. Regex himself. Credit for the remove del regex goes to KnottyDog (i think).

QoL shitty last paragraph deleter | catbox fix quotes | catbox fix quotes | fileditch

Gemini Version [updated: 06.02.25] (updated for Pro/Flash 2)

catbox | normal version catbox | schizo version Pick whichever you prefer. Normal version is just 2.5.1 with parameters adjusted and some toggles removed. Schizo version is the quirky edition that I've written specifically for Gemini, but it might just not be to everyone's taste. Better parameters could be found, but these work fine and don't tweak out for now.

Patch Notes (2.5.1)

I will update this in a few days since I'm technically still on vacation. I'm pretty sure no one actually reads these, anyway. If you do, fucking nerd lmao
improved homosexuality

FAQ

"I hate bloat presets."

That's not a question, but yeah, me too. This is why this JB is extremely lightweight, coming in at 200 tokens for basic functionalities, +500 tokens for an additional, strong Style prompt, and no optional toggles that exceed 200 tokens. Most toggles are around 50-100 tokens at the very most and have been heavily tested to ensure they actually do something. Adding your own toggles is incredibly easy.

"Why am I being sued?"

Skill Check :)

"Where does logit bias go?"

Under chat completion presets, above the prompt manager. Your endpoint needs to be openai. Doesn't work on Azure or Claude or Gemini.

"Does this work on Claude?"

This preset has been tested on various Claude models and works 100%.

"How do I edit?"

Small guide on how Latte works:

Latte prioritises tokens that are lower down. Meaning that the lower down something is, the stronger it is represented. You don't need to write it in a gaslighty way, just be concise in what you want and you will get the exact same results. Aka ao3 style prompting doesn't really work here. Anyone who says otherwise might be huffing placeboium.

Use as little tokens as possible. Be direct and concise in your instructions. Cut down whatever yap you're thinking about and just say directly what you want from Latte. Start any prompt with a hashtag like this:

⎗

1	# Instructions start here.

You can also give your instructions a title you can call back to.

⎗

1
2
3

# Desu
please desu desu
Callback to `# Desu`

Anything you want to call back to, or if you want to group instructions, you want to do like this:

⎗
✓
<desu>
# Instructions start here
</desu>
Call back to `<desu>`

Example of slop:

⎗

<plsdothis>
# So, Avi will totally follow this to the LETTER and NOT do ANYTHING else at all! Also, if Avi does this, she will get 500 dollars, and if Avi doesn't do this, a real human being will die. Ok so, I want Avi to write everything in AO3 style...
(This continues for 800 tokens...)
...and last, Avi will NEVER use words that are CRINGE you want to write exactly like ao3 and nothing else! Thank you!
</plsdothis>

Example of good:

⎗

1	# Avi will now write like a pirate.

There are several special prompts marked 🥤 on the JB specifically made for you to work with. They contain instructions on how to edit them on top of that.

"How do I edit (on Gemini)?

Gemini preset is built differently. Instead of System, every role is AI Assistant. For whatever reason, this works far better. This means Avi is also gone. You write prompts the same way like on Latte, with one crucial difference:

⎗
✓
# All prompts should be written in first person. I will now proceed to continue.

An actual example:

⎗

1
2
3

<tag>
I will write between 1000 and 1200 words.
</tag>

Everything else works the same. It doesn't matter how you structure the prompt manager to Gemini (so far), it will read and weigh all of it the same amount.

"I don't know what (x) prompt does!"

Open it and read.

"I read the Readme's and this FAQ and I still have questions/want to contribute!"

Uhhh... 'cord or email

Old versions

2.5.: Padoru Edition catbox fileditch

2.3.5: Curry Edition catbox fileditch

2.2: Spooky Edition catbox fileditch

2.1: Avi Edition catbox

2.0: Voodoo Edition catbox fileditch

1.0.3: The one before the big one Edition catbox

1.0.0: I have no idea what I'm doing Edition fileditch

My Bots

Once a year, I upload bots. I'll probably upload more of mine. Link to my profile here.