AI Prompt Studio

Learn prompting

Describe your idea. Copy your prompt. Choose a tool.

See how in 15 seconds

Your request and ready prompt

Your request

Customize request
Extra details
Task and template: Writing

A local keyword suggestion, not AI analysis. Review the task and template.

Writing suggestions (2)

Checks missing details and a few common typos. Not a complete grammar checker; no text is sent to a server.

  • Describe your intended result before building a prompt.

  • Specify your audience or listeners to tailor the output.

Ready prompt

Start with what you need

Editable examples — one click to start

A copyable prompt, not generated media. Results may need revisions.

Prompt options

Choose a tool for your request

Privacy and usage options

Fit and ease scores out of 5 are editorial estimates based on documented features, not independent quality benchmarks or outcome guarantees. Verify plans before paying. Reviewed:

Tailor your platform decision

Set your needs to reorder suggestions and explain trade-offs. These preferences stay in your browser session and are not sent to providers.

Provisional match for these preferences: Claude · Claude (plan dependent)

Compare all tools (7)
Provider and model comparison for current task
PickPlatform / modelOriginTask fit*Ease*AccessBilling
ClaudeClaude (plan dependent)Matched to your needsUS5/55/5Limited free + paidPlan limits
ChatGPTModel varies by planUS5/55/5Limited free + paidPlan limits
GeminiGemini (plan dependent)US4/55/5Limited free + paidPlan limits
DeepSeekDeepSeek Chat (current web model)China4/54/5Limited free + paidPlan limits
Le ChatMistralEurope4/54/5Limited free + paidPlan limits
OpenRouterFree models (:free, current router)US4/54/5Limited free + paidAPI tokens
OllamaLocal Qwen / GemmaDoes not match constraintsGlobal project4/52/5Self-host / openCompute / GPU
Advanced settings and API cost estimates
Model selection priority

Suggestions are guidance, not quality or availability guarantees. You can always select a model manually.

Task-based model recommendation

Suggested model: Google · Gemini 3.5 Flash-Lite

Lower-cost routine tasks; validate output quality

Choose Images or Video to find Seedream, Seedance, Wan, FLUX, Qwen, LTX and Hunyuan. Hosted and local GPU options differ; not every choice is free or runs in AICI.

Prompt text tokens (approx.)

0–0

Estimated cost (USD)

$0.00020

Text-only estimate using input and output token ceilings, not an invoice.

Official model pricing

Compare alternatives

Compare estimated tokens and model cost

Text token and cost planner

Illustrative estimate, not an exact provider count or invoice. Image/video generation fees are excluded.

Estimated input tokens

0–0

USD text cost at output ceiling

$0.00020–$0.00020

Estimated text costs by model
ModelEstimated USD cost
Google · Gemini 2.5 Flash-Lite ✓$0.00020–$0.00020
Google · Gemini 3.5 Flash-Lite$0.00125–$0.00125
Google · Gemini 2.5 Flash (legacy until Dec 2026)$0.00125–$0.00125
OpenAI · GPT-5.4 Mini$0.00225–$0.00225
Anthropic · Claude Haiku 4.5$0.00250–$0.00250
Anthropic · Claude Sonnet 4.5$0.00750–$0.00750

Reference prices as of 9 Nov 2026 may change. Actual token counts vary by model, language and context. Reasoning, tools, chat history and other fees may apply. Official pricing

Privacy and usage

Try your prompt on a provider

Building and copying here remain local. If text trials are enabled and you click Generate, your prompt is sent to the AI provider. Image and video generation are not available inside AICI yet.

More: campaigns, research and FAQs

AICI CAMPAIGN FACTORY

Turn your product idea into a usable marketing workflow

Describe your product once. Get six connected, copy-ready prompts for strategy, organic posts, ads, images, video and quality assurance — without sending your details to an AI provider.

Composed locally · No AI generation charges

1. Describe your campaign

Product, description and audience are required. Do not include secrets or sensitive personal information.

0/800

Customize voice and call to action (optional)

0 of 3 required fields complete

  • Product: at least 2 characters
  • Description: at least 12 characters
  • Audience: at least 2 characters

2. Your copy-ready workflow

0/6

Six connected tasks will appear here

  1. 1Seven-day strategy
  2. 2Three tailored posts
  3. 3Two ad variations
  4. 4Product image prompt
  5. 5Video storyboard and prompt
  6. 6Pre-publication review checklist

The factory generates prompts for use in external AI tools. It does not directly produce ads, images or videos.

Compare AI models and indicative cost

AICI · OPEN MODEL DISCOVERY

Explore open GitHub image, video and audio projects

These are real code/model projects, not offers of unlimited free tokens or free GPUs. Stars and forks reflect community attention—not visual quality, model downloads or customer demand.

Research snapshot: . GitHub metrics: . Metrics are not live. Review model-weight licenses and local hardware requirements before use.

Hugging Face model download signals

Reported 30-day downloads from indexed official model cards during research. Some cached pages may lag. These count specific model repositories—not unique users, generated-video downloads or free hosted credits. Research snapshot:

Qwen-Image — ComfyUI packaged weights

Images

2,647,004

reported downloads / 30 days (cached snapshot)

Counts the Comfy-Org model package, not QwenLM code downloads or unique users. Verify the weights license.

Verify current downloads and model terms

LTX-2.5 — audio/video model weights

Video

1,607,815

reported downloads / 30 days (cached snapshot)

Weights use a community license, not unrestricted Apache rights. Downloading does not include free GPU inference.

Verify current downloads and model terms

HunyuanVideo 1.5 — ComfyUI packaged weights

Video

741,891

reported downloads / 30 days (cached snapshot)

Counts this ComfyUI model package, not all Tencent variants; GPU and Hunyuan license requirements apply.

Verify current downloads and model terms

Qwen-Image 2.1 — research-licensed model

Images

261,646

reported downloads / 30 days (cached snapshot)

The model uses a Qwen Research License, not unrestricted commercial Apache terms.

Verify current downloads and model terms

Showing 11 of 11 researched projects

Stable Diffusion WebUI

Images

AUTOMATIC1111/stable-diffusion-webui

Popular image generation and editing UI for locally run Stable Diffusion models

Stars
165,287
Forks
31,155
Last push
2026-03-02

Repository license: AGPL-3.0

Model-weight license review: required for each model

Very popular, but repository push activity is slower; review extension support and model licenses.

View repository and models

ComfyUI

Multimodal

Comfy-Org/ComfyUI

Local workflows for images, video and audio; supports many different models

Stars
146,366
Forks
17,433
Last push
2026-11-11

Repository license: GPL-3.0

Model-weight license review: required for each model

Optional paid nodes cost money and model licenses are separate. Support for a model does not prove that particular model runs in 4GB VRAM.

View repository and models

Hugging Face Diffusers

Multimodal

huggingface/diffusers

Developer inference library for many image, video and audio pipelines

Stars
34,910
Forks
7,478
Last push
2026-11-11

Repository license: Apache-2.0

Model-weight license review: required for each model

Not a provider of free credits or free GPUs; models and compute must be supplied.

View repository and models

InvokeAI

Images

invoke-ai/InvokeAI

Image generation, inpainting, canvas editing and repeatable creative workflows

Stars
29,070
Forks
3,103
Last push
2026-11-11

Repository license: GitHub: Apache-2.0 (+ other project licenses)

Model-weight license review: required for each model

Some supported models require paid API access; verify component and weight licenses.

View repository and models

FLUX.1 schnell

Images

black-forest-labs/flux

Run the FLUX.1 schnell image model locally under its Apache-2.0 license

Stars
26,058
Forks
1,933
Last push
2025-07-31

Repository license: FLUX.1 schnell: Apache-2.0

Model-weight license review: required for each model

Do not confuse schnell with FLUX.1 dev, whose weights have different non-commercial terms. Requires GPU.

View repository and models

Wan 2.2

Video

Wan-Video/Wan2.2

Local text-to-video and image-to-video with model-size-specific requirements

Stars
18,552
Forks
2,464
Last push
2026-09-21

Repository license: Apache-2.0

Model-weight license review: required for each model

Large checkpoints and compute-intensive inference; verify weights and commercial terms.

View repository and models

LTX-2 family

Video

Lightricks/LTX-2

Synchronized audio and video generation with the supported LTX checkpoints

Stars
10,194
Forks
1,622
Last push
2026-11-05

Repository license: Apache code + separate LTX model terms

Model-weight license review: required; commercial restrictions

Python-code license does not license weights. LTX-2.5 has separate community and commercial revenue terms.

View repository and models

Qwen-Image

Images

QwenLM/Qwen-Image

Open-weight image model including text rendering use cases

Stars
8,475
Forks
569
Last push
2026-02-10

Repository license: Apache-2.0 (repo)

Model-weight license review: required for each model

Review the model card, commercial terms and VRAM requirements; no free cloud GPU is included.

View repository and models

WanVideo ComfyUI wrapper

Video

Kijai/ComfyUI-WanVideoWrapper

ComfyUI nodes and workflows for Wan video checkpoints

Stars
6,813
Forks
714
Last push
2026-05-24

Repository license: Apache-2.0 (wrapper)

Model-weight license review: required for each model

An integration wrapper, not free GPU time. Review third-party nodes before installation.

View repository and models

ACE-Step

Music

ace-step/ACE-Step

Generate original music locally using ACE-Step, with compute rather than hosted credits

Stars
5,051
Forks
664
Last push
2026-02-15

Repository license: Apache-2.0 (repo)

Model-weight license review: required for each model

Verify weights, dataset and output rights; do not assume unrestricted commercial licensing.

View repository and models

HunyuanVideo 1.5

Video

Tencent-Hunyuan/HunyuanVideo-1.5

480p/720p video model; official minimum 14GB GPU VRAM with model offloading

Stars
4,598
Forks
238
Last push
2026-04-10

Repository license: Tencent Hunyuan community terms

Model-weight license review: required; commercial restrictions

Tencent Hunyuan community licensing is not unrestricted Apache weights; geographic and commercial restrictions may apply.

View repository and models

True model-download demand requires separately verified Hugging Face or GitHub release download metrics. It cannot be inferred from stars; generated-content demand is not measured by these repositories.

When should you use the AI campaign planner?

Launch a product or service

Start with verified features, an audience and a call to action. Prepare connected strategy, posts and ad prompts, then verify prices and claims before publishing.

Plan social media content

Choose Instagram, Facebook, LinkedIn or TikTok. Use the prompt kit to brief the writing, design and video tools appropriate for each task.

Review a campaign before execution

Use the quality review prompt to check clarity, facts and rights. Generation and publishing happen in the tools you choose; neither happens automatically here.

Start planning your campaign

Using AI to help write code?

A useful prompt does not replace a security review. For production business applications, review authorization, secrets and APIs before release.

Explore secure code review

From an idea to useful output

1 · Define the goal

State the goal, audience, platform and references. Keep private data and access keys out of prompts.

2 · Set constraints

Specify length, tone, dimensions, video duration and what must remain in an image. Ask for one question only when a critical detail is missing.

3 · Review output

Check facts, rights, logo accuracy and visual similarity. Revise a specific part instead of regenerating everything.

Future plans — subscriptions are not active

AICI AI Prompt Lab

Daily trials are proposed for text only. Images and video need separate monthly credits and firm cost caps. Indicative prices require cost and payment validation.

Starter

Free

Up to 5 daily text runs after launch. Unlimited compose/copy. Media prompts only.

Advanced

EGP 200 / month (proposal)

Up to 20 daily text runs, smart routing, advanced tools; images through a separately capped pool.

Creator Pro

EGP 599 / month (proposal)

Up to 50 daily text runs, social presets, capped images, video via credits or pay-as-you-go.

No AI credits, generation service, or subscription checkout is enabled yet. These are not unlimited-use entitlements.

Open learning references

Learn from public methodologies and author original templates instead of republishing other creators' work. Check licenses before reuse.

DAIR.AI Prompt Engineering Guide (GitHub · MIT)

AI prompt and cost FAQs

Can I use these prompts with ChatGPT, Claude and Gemini?

Yes. Templates are portable text, but each model has different capabilities and limitations. Review results and adjust where needed.

Is the token estimate the actual API bill?

No. Estimates are planning tools. Real billing can include output tokens, tools, context and separate media fees.

Does the studio generate images, video, audio or music?

No. Studio creates copyable prompts and shows indicative model rates. Any enabled in-site trial is text-only and subject to daily limits.

How to reduce cost

Start with a specific request and short output. Validate the idea with an economical text model before generating images or video. Prompt tokens alone do not predict media costs; output quality and video duration vary by provider.