Yodayo AI

Signing up and landing on ten tools at once

Account creation asks for no payment details, and there is no onboarding sequence afterwards.

The landing screen is a grid of ten separate products: Chat in Tavern, Posts, Image Gen, Chat with Yoda-chan, Video Gen, Music Gen, Voice, Model Hub, Image Presets and Lora Training.

The balance counter in the top right is the only number that matters on this page, and nothing on the grid explains what any tile costs to run.

YoBeans and the two day clock

The currency is called YoBeans on every screen of the Yodayo interface. Several published reviews still call it Mochi, which is the name used on the Moescape side of the same company.

The rewards page splits your balance into buckets with different rules attached to each one. New User YoBeans carry no expiry. Daily Free YoBeans do.

The activity log states it twice: 50 claimed on August 24, expiring 12:26 PM on August 26. That is two days, not the thirty day window several review sites report, and Yodayo's own help documentation says 24 hours.

Spend order does work in your favor. My daily allowance fell from 50 to 34 while the signup 100 sat untouched, so the perishable credits go first.

What each action actually costs

There is no separate chat plan or image plan. Every feature draws from the same balance, so a long roleplay session is artwork you did not generate.

These are the prices I logged on a new account:

A day's 50 credits covers eight images at the default resolution. The same 50 covers twelve chat messages.

First generation: strong style, loose instruction following

I ran one fixed control prompt with a negative prompt attached so both could be reused in later tests.

The negative field came preloaded with the usual defaults: worst quality, low quality, blurry, extra fingers, bad anatomy, watermark, signature, text, jpeg artifacts.

Note the Generate button. It prints its price before you commit, which is more honest than most credit-based generators manage.

The canvas default nobody tells you about

The generator opens at 512x512 and stays there. The size control sits in the Settings panel on the right, collapsed under Canvas Size, and nothing prompts you to open it.

A community image I inspected later in the same session ran at 1280x960, close to five times the pixel area.

New accounts therefore generate at the cheapest and lowest setting by default, and judge the platform's output quality on that basis.

Where the model breaks: the hands test

The second test targeted the classic failure case, with the negative prompt field cleared so nothing was suppressing bad anatomy.

Six YoBeans later, here is what came back.

Image-to-Prompt only speaks anime

Image-to-Prompt reads an uploaded picture and writes the prompt that would recreate it, then offers to push that prompt into the generator.

I fed it a stock photograph of six people around a meeting table, deliberately outside the platform's home territory.

The community feed is the real documentation

Every public image on the platform publishes its full generation data, and this is the single strongest reason to use Yodayo over a standalone generator.

Model, Spell, sampling method, steps, CFG scale, seed, canvas size and the complete prompt are all visible. A button below loads that entire configuration into your own generator.

Tavern: deep settings, most of them switched off

Character creation offers two routes. YoPilot writes a character from three fields: name, description and gender. Manual entry gives full control and accepts a JSON upload.

That JSON import matters more than it looks, because character cards built on other roleplay platforms transfer across intact.

Inside a chat, the settings panel runs deeper than I expected: System Prompt, Presets, Persona, Memory, Scenario and Lorebook, with the language model selector at the top.