Celestia — AI assistant for photos, videos and music | MindlyFlow
MindlyFlow AI assistant

Celestia —
an AI assistant for your ideas

Describe what you want to create. Celestia understands the idea, writes the prompt, picks the format and builds a plan with a price. Tokens are charged only after you press “Create”.

Celestia AI assistant · online Dialog

Make a video: a rainy night city, neon reflected in puddles, the camera drifts slowly over the street

Here is the plan: 16:9, 5 seconds, neon reflections and rain sound. Check the settings and press “Create”.

Mindly Video 16:9 · 5 sec · 10 tokens Create
Tokens are charged only after you launch
Done
  • Gets the idea in a sentence
  • Writes prompts for photos, videos and music
  • Remembers frames and chat context
See examples
Chat is free — you pay only for generations
Free chat
Messages don't use tokens
Plan before charging
Price and settings upfront
Keeps the context
Frames, edits and history
All tools
Photo, video, music, animation
What Celestia can do

More than a chat — a working assistant

She takes an idea to a ready plan: from the first message to the “Create” button.

Gets the idea right away

Understands the idea and asks one precise question only when needed.

Builds the prompt

Describes style, light, camera and sound — with no generic filler.

Keeps the context

Uses chat frames, including the first one, as a reference.

Shows the price

Model, format and cost — before any tokens are charged.

Examples

From a request — to the result

Four scenarios: the dialogue, the plan with a price, and the finished piece. Any of them can be repeated in the chat.

Celestia AI assistant · online

A portrait in a coffee shop: warm evening light, 85 mm, soft bokeh

Here is the plan: a 1:1 frame, FullHD. I added light, lens and interior details.

Mindly Image 1:1 · FullHD · 2 tokens Create

Prompt Photoreal, a portrait in a coffee shop, warm evening light, 85 mm, soft bokeh

I can make 2–4 variants of the frame — each is paid separately.

Tokens are charged only after you launch

Portrait in a cafe
Portrait in a cafe 1:1 · FullHD · 2 tokens Repeat in chat

The example results were made in MindlyFlow. The plan and price depend on your settings.

Tools

One chat — every generation

Celestia knows which tool you need and builds a plan for the task. Prices come from the live price list.

Photo creation TEXT TO IMAGE

Photo

A frame from a description: light, lens, details

from 2 tokens
Video creation VIDEO

Video

A clip with sound: 4–15 seconds

from 10 tokens
Photo revival HOT

Animation

Motion and expression for a finished frame

from 10 tokens
NEW

Music

A song or instrumental from an idea

from 3 tokens
Photo Editor EDITOR

Photo editing

Change a finished frame from a description

from 3 tokens
DeepFake DEEPFAKE

Deepfake

Face Swap on Photos and Videos

from 1 token
Prompts

Celestia writes prompts like a film director

You describe the idea — she adds what shapes the result: composition, light, camera movement and sound.

Photoreal, a medium shot, a boat drifts across a lake at sunrise, the camera slowly pushes in, mist over the water, warm rim light, water lapping and morning birds.

Frame and style — photo, film, anime, 3D
Subject and action — who is in the frame and what they do
Camera — how the imaginary camera moves
Light and atmosphere — time of day, weather, details
Sound — what you hear in the clip

Specifics over adjectives

Not “a beautiful shot”, but “mist over a lake, 85 mm, soft rim light”.

Motion and sound

For video: what happens in the frame, how the camera moves and what you hear.

Storyboard

If the scene has several actions, use shots with timecodes, like in film.

Prompt language

Photos and videos use English, music uses your language. Celestia translates for you.

Open chat
Context and memory

Remembers frames and you

Celestia sees the frames in the chat: edit the latest one, use the first as a face reference, or build a series with the same character.

  • Up to 5 references in one plan
  • Edit the current frame without starting over
  • The “Celestia memory” note: style, light, language
  • Chat history is stored in your account
Open chat
Celestia AI assistant · online Context
1 2 3 4 chat frames

Use the first frame for the face and the second for the light, make a 9:16 video, 6 seconds

Got it: reference 1 is the character, reference 2 is the light. The plan: a 9:16 video, 6 seconds.

Mindly Video 9:16 · 6 sec · 10 tokens Create
Memory 35 mm, warm light, I write in Russian
How It Works

From idea to result — in three steps

Celestia handles the prompt and settings, while the launch decision stays with you.

1

Describe the idea

One sentence or a ready prompt — Celestia asks for details only when needed.

2

Review the plan

Format, duration, references and cost are all visible before launch.

3

Click “Create”

Tokens are charged only after confirmation. The result appears in the chat and in “My files”.

Use cases

What it is for

When describing an idea is easier than figuring out tools and prompts.

You do not know where to start

Celestia suggests a tool and builds your first plan.

An idea in words

Turns “I want it beautiful” into a concrete prompt with light and lens.

A series with one character

Builds a character sheet and new angles from your photo.

A storyboard for video

Writes shots with timecodes and camera movement.

Editing a finished frame

Replaces the background, adds rain or changes the light from a description.

A song on demand

Comes up with the idea, lyrics and mood of the track — with or without vocals.

Social Media Content

Ideas, scripts and frames for Reels, Shorts and posts.

A plan for a client

Shows the scope, format and price before launch — easy to approve.

Start for free

Have an idea? Describe it to Celestia

Open the chat and describe what you want to create. The first plan with a price will appear in seconds.

  • Chat without tokens
  • Plan and price before launch
  • Chat history in your account
  • Tokens never expire

She is MindlyFlow's AI assistant. She helps shape an idea, write a prompt, pick a tool and build a generation plan with a price.

Yes, chat messages don't use tokens. Tokens are only needed to start a generation, and the price is always shown in the plan before confirmation.

She knows MindlyFlow's tools and prices, sees the images in the chat and builds a ready plan: mode, prompt, settings and cost. Nothing starts without your confirmation.

Yes. Upload a JPG, PNG or WebP — Celestia will use it as a reference for editing, animation or video.

Chat history is stored in your account. In the current conversation she remembers the context and images, and there is a “Celestia's memory” note for lasting preferences.

No. She builds the plan and shows the price, and you confirm the launch with the “Create” button. Tokens are never charged without your consent.

Photos, videos, animation, music and photo editing. For Deepfake she helps with the idea and prompt, while the launch stays in a separate mode.

Tokens are refunded automatically. Talking to Celestia does not spend tokens — they are only needed to launch a generation.

An AI assistant for photos, videos and music — how Celestia works

What Celestia is

Celestia is an AI assistant inside MindlyFlow. She understands an idea in Russian or English, helps with the prompt and builds a plan: tool, format, duration and price. She is not a separate model, but a guide to the service tools.

How It Works

A request becomes a structured prompt: subject, scene, light, camera and sound. For video, Celestia can write shots with timecodes. The plan shows the cost before launch, and tokens are charged only after you confirm.

Who it is for

Beginners — so they do not have to learn prompts and settings. Authors and marketers — to get series of frames, clips and music quickly. Designers and production teams — to check the idea and the price before launch.