AI & LLM Tools

Token counters, LLM cost calculators, and AI model tools.

AI Token Counter

Count the estimated tokens in any text or prompt.

LLM Token Cost Calculator

Estimate the cost of LLM token usage across models.

Multilingual Token Counter

Estimate tokens for text in any writing system.

Tokenizer Visualizer

See how text splits into approximate tokens.

Context Window Calculator

Check how much of a model's context window is used.

Prompt Cost Calculator

Estimate the token cost of a single prompt.

OpenAI Cost Calculator

Estimate OpenAI API costs for GPT models.

Claude Cost Calculator

Estimate Anthropic Claude API costs with caching.

Gemini Cost Calculator

Estimate Google Gemini API costs.

Prompt Length Checker

Verify a prompt fits a model's context window.

Token Efficiency Calculator

Measure how many tokens a rewritten prompt saves.

RAG Chunking Calculator

Plan chunk size and overlap for RAG pipelines.

Embedding Cost Calculator

Estimate embedding API costs for your corpus.

Prompt Injection Checker

Scan prompts for common injection and jailbreak patterns.

AI Model Comparison

Compare context windows and pricing across providers.

About AI & LLM Tools

Free online AI tools for developers, prompt engineers, and anyone building on large language models. Count tokens and estimate the cost of prompts with OpenAI, Anthropic Claude, and Google Gemini, explore tokenizer behavior, check context window fit, plan RAG chunking, estimate embedding costs, scan prompts for injection patterns, and compare AI models side by side. Token counting uses a transparent approximation tuned per writing system, and cost estimates use the published per-million-token prices of each model. Everything runs directly in your browser, so your prompts and text are processed locally and never uploaded to a server.

Frequently Asked Questions

Token counting in this toolkit is an approximation. Real tokenizers such as OpenAI's cl100k_base are more complex and may differ by a few percent, especially for code, long words, and mixed scripts. Use the estimate as a planning figure rather than a billing-precise count.

The estimates use the per-million-token prices published by OpenAI, Anthropic, and Google at the time the models were added. Model pricing and availability change frequently, so always confirm the current prices on the provider's pricing page before budgeting.

Tokenizers split text by predicting token boundaries. Latin script roughly averages 4 characters per token, while scripts such as CJK, Arabic, and Thai commonly average around one token per character, so multilingual text estimates are weighted by script.

No. All token counting, cost estimation, and prompt analysis are computed locally in your browser using the text you paste. Your data does not need to be uploaded to a server.

The context window is the maximum number of tokens a model can process in a single conversation, including both the system prompt, the conversation history, and the generated output. Exceeding it causes errors or truncation.

An embedding model converts text into a vector of numbers. Billing is typically per million tokens of text you send for embedding, so the embedding cost depends on the total number of tokens across all documents or chunks.