Free AI & Prompt Tools
Prompt tools for working with AI assistants and image generators: build a structured prompt from a role, context, task, tone, format and constraints; draft a system prompt from toggled sections and templates; build an image prompt with style, lighting, camera and mood controls; and estimate how many tokens a text uses against a model's context window. You copy the result into the assistant you use.
None of these tools calls an AI model, so there is no API key to enter, no usage limit and nothing sent to a provider. They organise your thinking and your text. Token counts are approximations, because each model family counts tokens in its own way.
4 tools available
Most Popular
Build structured AI prompts with role, context, task, tone, format, and constraints. Works with ChatGPT, Claude, Gemini, and any LLM.
Try it free →Token CounterCount approximate tokens in your text for GPT-4, Claude, Gemini, and Llama. See context window usage and estimated API cost per model.
Try it free →System Prompt BuilderBuild structured system prompts for ChatGPT, Claude, and Gemini visually. Toggle sections, load templates, and copy the final prompt.
Try it free →Which tool should I use?
- A clear, structured request: AI Prompt Builder — role, context, task, tone, format and constraints for ChatGPT, Claude, Gemini and others.
- Standing instructions for an assistant: System Prompt Builder — toggleable sections and templates for a system prompt.
- A picture from a text-to-image model: AI Image Prompt Builder — style, lighting, camera and mood for Midjourney, DALL-E 3, Stable Diffusion and Firefly.
- Will this fit in the context window?: Token Counter — approximate tokens for GPT-4, Claude, Gemini and Llama, with context use and an estimated API cost.
All AI & Prompt Tools (4)
Build structured AI prompts with role, context, task, tone, format, and constraints. Works with ChatGPT, Claude, Gemini, and any LLM.
Open tool →Token CounterCount approximate tokens in your text for GPT-4, Claude, Gemini, and Llama. See context window usage and estimated API cost per model.
Open tool →System Prompt BuilderBuild structured system prompts for ChatGPT, Claude, and Gemini visually. Toggle sections, load templates, and copy the final prompt.
Open tool →AI Image Prompt BuilderGenerate AI image prompts for Midjourney, DALL·E 3, Stable Diffusion, and Firefly. Choose style, lighting, camera, mood, and aspect ratio.
Open tool →Frequently Asked Questions
- Do I need an API key?
- No. These tools build and count text; they do not send anything to an AI service. You copy the finished prompt into whichever assistant or image generator you use.
- What makes a prompt work better?
- Say who the assistant should act as, give the background it needs, state the task in one clear sentence, and describe the format you want back, with any limits such as length or things to avoid. The prompt builder has a field for each of these, so nothing important is left out.
- What is the difference between a system prompt and a normal prompt?
- A system prompt sets standing instructions for how the assistant should behave across a whole conversation, such as its role, tone and rules. A normal prompt is the request you make in a single message. The system prompt builder is for the first, the prompt builder for the second.
- How accurate is the token counter?
- It gives an approximation for GPT-4, Claude, Gemini and Llama style tokenisers. Exact counts differ by model and version, so leave some margin when you are close to a context limit, and treat the cost estimate as a guide.
- Is my prompt stored?
- No. Prompts are built and counted in your browser and are not sent to Stax or saved.