AI Utilities Tools
Estimate tokens and API cost for common chat models from published tokenizers and price lists. We do not call those APIs with your prompt.
7 tools · 100% browser-based · No uploads · No signup
About these AI Utilities tools
AI utilities estimate tokens, visualize token boundaries, plan context windows, trim prompts, and ballpark API cost for common chat and embedding models. They run locally with published tokenizers and price tables. They do not call OpenAI, Anthropic, or Google on your behalf.
Token counts are estimates of what a given tokenizer would emit. Providers can change tokenizers and prices. Always confirm on the vendor's pricing page before you commit budget. Cost compare is a worksheet, not a contract.
Tools in this category
Token Counter
AI token counter online. Count tokens for GPT-5.6, Claude 5, Gemini 3.x, Grok 4.6, and DeepSeek V4. Character count and cost estimation in this browser.
Token Visualizer
Shows how this encoding splits text: leading spaces, camelCase, CJK. Blocks are OpenAI BPE; Claude counts are scaled. Not a generic token counter.
Context Estimator
Window math: system + history + prompt + output must fit. Remaining vs max_tokens, not a token counter online.
Prompt Trimmer
AI prompt trimmer. Trim text to fit token limits with sentence-aware and paragraph-aware modes. Nothing is uploaded.
Embedding Estimator
AI embedding calculator. Estimate tokens, dimensions, and vector DB size for text-embedding-4-large, Gemini Embedding 2, and Cohere.
Cost Estimator
AI API cost calculator. Estimate GPT-5.6, Claude 5, Gemini 3.7 Flash, and DeepSeek V4 bills from published rates (updated 20 Aug 2026).
Cost Compare
Compare AI API costs. Monthly pricing across GPT-5.6 Sol/Terra/Luna, Claude Fable 5 / Opus 5 / Sonnet 5, Gemini 3.x, Grok 4.6, and DeepSeek V4.
How these tools run in your browser
Tokenizers (where bundled) run in the browser. Some models use a close sibling tokenizer when the exact one is too large to ship. Prices live in our source and are dated. Context estimator adds system, tools, and history the way typical chat APIs bill, which is still an approximation of each provider's real billing rules.
Common Use Cases
- See whether a prompt plus history fits an 8k, 32k, or 128k window
- Trim logs out of a prompt without cutting mid-sentence
- Compare input/output prices across models for a daily volume
- Estimate embedding dimensions and storage for a vector DB
- Explain to a stakeholder why "characters" are not "tokens"
Common Mistakes
Frequently Asked Questions
No. Counting and trimming happen on this page. There is no chat completion.
Different tokenizer builds, special tokens, and chat templates. Use the vendor playground as the source of truth for billing disputes.
We review provider pages at least monthly. See the editorial policy. Check the last-updated date on the cost tools.