PocketKit
Free AI and LLM tools
Use token counting, model pricing, prompt generation and tool-schema conversion utilities for Claude, GPT, Gemini and DeepSeek workflows.
What can you do with these tools?
AI and LLM utilities focus on repeatable development tasks rather than replacing the models themselves: estimate tokens and API cost, compare model pricing, structure system prompts and convert tool definitions across Claude, OpenAI and Gemini formats. Pricing and token results explain their date or estimation method; verify important production decisions with provider documentation.
Start here
Related tools
AI Token Counter & Cost Estimator
Estimate token counts locally and compare API costs across Claude, GPT, Gemini and DeepSeek models
GPT Token Counter
Count GPT-4o/4/3.5 tokens exactly with OpenAI's official BPE (o200k/cl100k) locally and estimate cost
LLM API Pricing Comparison
Compare input/output/cached prices, context windows and blended cost across Claude, GPT, Gemini and DeepSeek
AI Prompt & System Prompt Generator
Structure and optimize system prompts for DeepSeek, Claude, and ChatGPT with code review, translation and content templates
Tool & Function Schema Converter
Convert between Claude tool, OpenAI function, Gemini functionDeclaration and JSON Schema — with strict mode
AI Image Ratio & Resolution Calculator
Calculate exact pixel resolution, aspect ratios and --ar parameters for Midjourney, Flux.1, and SDXL image generators
AI Background Remover
One-click AI cutout in your browser — no upload
Decision guide
LLM tasks and result confidence
| Task | Method | Source of truth |
|---|---|---|
| Token budgeting | Matching tokenizer or coefficient estimate | AI Token Counter & Cost Estimator |
| API cost | Token count times model price | LLM API Pricing Comparison |
| Prompt structuring | Separate role, constraints, input and output | AI Prompt & System Prompt Generator |
| Tool-definition migration | Convert schema fields and review manually | Tool & Function Schema Converter |
FAQ
Frequently asked questions
Are token counts exact?
Some models use a matching tokenizer, while others may use estimation coefficients. The tool page explains its method; use API-reported usage for billing reconciliation.
How often is model pricing updated?
The pricing page identifies its review date and reference sources. Providers can change prices at any time, so check their official pricing pages before final budgeting.
Are prompts and text sent to an AI model?
Prompt structuring, token estimation and format conversion do not require calling an LLM API. Any feature that needs a network request explains it on its page.