AI Tools
AI Tools gathers the practical utilities people need once they move past typing questions into a chat box: writing prompts that get consistent answers, counting and cutting tokens, estimating what an API bill will be, checking a fine-tuning dataset, sizing GPUs, planning a team of agents led by a chief-of-staff agent, connecting MCP servers to Claude, Cursor or VS Code, building an MCP server for your own app, tracing and auditing what agents do, and making a website readable by AI assistants and AI search.
30 free tools · no sign-up · no ads
Build and lint prompts, count and trim tokens, and estimate LLM API spend across models.
AI Prompt Builder
Build structured prompts with role, context, examples and output formatPrompt Quality Checker
Score a prompt and get fixes for vague or missing instructionsLLM Token Counter
Count tokens for GPT, Claude, Gemini and Llama, with costPrompt Token Reducer
Cut tokens from prompts and context without losing meaningLLM API Cost Calculator
Monthly API cost by model, with caching and batch discounts
Check fine-tuning data, size GPUs, estimate training compute and score model outputs.
Fine-Tuning Dataset Validator
Check and convert JSONL chat datasets before fine-tuningLLM GPU Memory Calculator
VRAM for inference, full fine-tuning, LoRA and QLoRALLM Training Compute Calculator
FLOPs, Chinchilla tokens, GPU hours and cost for pre-trainingLLM Eval Scorer
Score model outputs against expected answers with exact match, F1 and ROUGE
Plan agent teams, write agent prompts and tool schemas, and configure or build MCP servers.
Multi-Agent System Designer
Plan a chief-of-staff agent and specialist agents for a goalAgent System Prompt Generator
Write a system prompt with tools, guardrails and stop rulesAI Tool Schema Builder
Build function-calling tool definitions for OpenAI, Claude and MCPMCP Config Generator
Connect MCP servers to Claude, Cursor, VS Code and moreMCP Server Generator
Scaffold an MCP server in TypeScript or Python from tool definitionsOpenAPI to MCP Converter
Turn an OpenAPI spec into MCP tools and a server
Trace and audit what AI systems do, and make your site readable to AI assistants and search.
LLM Tracing Setup Generator
OpenTelemetry tracing code for LLM and agent appsAI Audit Log Designer
Design audit records for AI decisions and agent actionsllms.txt Generator
Create an llms.txt file so AI assistants can read your siteAI Crawler robots.txt Generator
Allow or block GPTBot, ClaudeBot, PerplexityBot and other AI crawlersAI Visibility Checker
Check what AI search and assistants can read on a page
Fill-in prompt templates for ten professions, with privacy checks and output review lists.
AI Prompts for Doctors
Ready prompts for notes, patient letters and literature, with privacy checksAI Prompts for Nurses
Prompts for handoffs, care plans and patient teachingAI Prompts for Civil Engineers
Prompts for RFIs, specs, site reports and calculation checksAI Prompts for QA Engineers
Prompts for test cases, bug reports and automation scriptsAI Prompts for Insurance Agents
Prompts for coverage explainers, renewals and client follow-upsAI Prompts for Front Desk Staff
Prompts for scheduling, replies, phone scripts and check-inAI Prompts for Executive Assistants
Prompts for briefings, inbox triage, travel and meeting notesAI Prompts for Restaurant Operators
Prompts for menus, reviews, staffing and food costAI Prompts for Marketers
Prompts for briefs, campaigns, SEO content and ad copyAI Prompts for Sales Teams
Prompts for prospecting, discovery, objections and follow-ups
Which tool do I need?
Get better, more consistent answers from a chatbot
Structure the prompt with role, context, examples and an output format, then lint it for vague or missing instructions.
Cut the cost of an LLM feature
Count tokens, remove the ones that add nothing, and compare monthly cost across models with caching and batch discounts.
LLM Token CounterPrompt Token ReducerLLM API Cost Calculator
Prepare to fine-tune or self-host a model
Validate the JSONL dataset, then work out how much GPU memory inference or LoRA training needs.
Have several agents work toward one goal
Design a chief-of-staff agent with specialist agents, budgets and approval points, and write each agent's instructions.
Let AI assistants use your app or service
Generate an MCP server from your tools or OpenAPI spec, then the config that connects it to each AI client.
MCP Server GeneratorOpenAPI to MCP ConverterMCP Config Generator
Make your site show up in AI answers
Write llms.txt, decide which AI crawlers to allow, and check what a page exposes to them.
llms.txt GeneratorAI Crawler robots.txt GeneratorAI Visibility Checker
About AI Tools
None of these tools call a language model, and none need an API key. They build, check and calculate in your browser, so you can paste a confidential prompt, a training file or an internal API spec without it leaving the tab. You take the result (a prompt, a config file, generated code, a cost estimate) to whichever model or platform you use. The profession kits apply the same idea to daily work for doctors, nurses, engineers, testers, insurance agents, front desk staff, executive assistants, restaurant operators, marketers and sales teams, with privacy checks before anything is copied.
Each tool links to a longer guide that explains the method behind it, such as why the Chinchilla rule suggests about 20 training tokens per parameter, how prompt caching changes cost, or what llms.txt is and is not.
Frequently asked questions
Do the AI tools send my prompts or files to an AI model?
No. They are generators, checkers and calculators that run in your browser. Nothing you type or upload is sent to RedBison or to a model provider. You copy the output into the AI product you already use.
Do I need an API key or an account to use them?
No. There is no sign-up and no key. Tools that would normally call an API, such as the token counter, ship the tokenizer with the page instead.
Which AI models do these tools support?
The prompt, agent and MCP tools are model-agnostic and produce text or config for ChatGPT, Claude, Gemini, Copilot and open models. The token counter is exact for OpenAI tokenizers and gives labelled estimates for others, and the cost calculator lists each provider's published prices with the date they were checked.
Are the profession prompt kits safe to use with patient or client data?
The kits warn before you copy and can mask obvious identifiers such as phone numbers, emails and record numbers in pasted text. Whether a given AI service may receive personal or health data depends on your organisation's agreements and the law that applies, so follow your employer's policy.
Can these tools tell me if my website appears in ChatGPT or Perplexity answers?
No tool can promise that. The visibility checker shows whether a page gives AI crawlers readable server-rendered content, structured data and clear answers, and whether robots.txt blocks them, which are the parts you control.