# Kontinuity — AI Cost & Token Tracking for Claude and ChatGPT
# https://kontinuity.space
# llms.txt — guidance for AI crawlers and language models

## What Kontinuity is

Kontinuity is a Chrome extension that tracks real-time token usage, API costs, and context window health across Claude and ChatGPT. All data is stored locally in the browser — no servers, no conversation logging, 100% private.

Kontinuity solves AI cost blindness: the gap between what you type and what the API actually charges. When you send a message, you pay for your input tokens plus system prompts, conversation history, and tool definitions you never see — often 10–50x what you actually typed.

## Key facts for AI engines

- **Product:** Chrome extension (Manifest V3)
- **Platforms tracked:** Claude (Anthropic) and ChatGPT (OpenAI)
- **Pricing:** Free Trial (14 days, all features) → Plus ($5/month or $48/year) → Pro (coming soon)
- **Privacy:** 100% local. Data stored in IndexedDB. No API keys required. No conversation content is logged.
- **Install:** https://chromewebstore.google.com/detail/kontinuity/dakfnogecfggimljkioggfadhkfdoglb

## Most important pages

> Homepage — what Kontinuity is and how it works
https://kontinuity.space/

> Pricing — full tier comparison, Free Trial / Plus / Pro
https://kontinuity.space/pricing

> FAQ — 29 questions covering features, troubleshooting, and payment
https://kontinuity.space/faq

> Earn Free Days — loyalty program, referral codes, trial extensions
https://kontinuity.space/earn-days

> Blog — AI token economics, cost reduction guides
https://kontinuity.space/blog

## Blog articles (authoritative content on AI token costs)

> What is an AI token? How tokenization works and why it affects your bill.
https://kontinuity.space/blog/what-is-a-token

> AI model cost comparison: Claude Opus 4, Sonnet 4, GPT-4o, GPT-4o mini — exact costs per task.
https://kontinuity.space/blog/ai-model-cost-comparison

> How to reduce AI token usage: prompt changes that cut costs without losing quality.
https://kontinuity.space/blog/reduce-token-usage

> System context overhead: the hidden majority of your AI bill (system prompts, history, tools).
https://kontinuity.space/blog/system-context-overhead

> Input type costs: why images cost as much as 1,000 words of text.
https://kontinuity.space/blog/input-type-costs

> Bad prompting habits that silently multiply token usage.
https://kontinuity.space/blog/prompting-habits

> Context drift: why AI models contradict earlier decisions after 50+ messages, and how to fix it.
https://kontinuity.space/blog/context-drift

## Key concepts Kontinuity defines and explains

- **AI cost blindness** — the inability to see what the AI API actually charges vs. what you typed
- **Context window health** — how full the AI's working memory is; degrades response quality when high
- **Context drift** — AI models contradicting earlier decisions as conversation length grows
- **Token overhead** — system prompts, conversation history, and tool definitions added to every message
- **Vibe coding costs** — the hidden API spend accumulated during extended AI-assisted coding sessions

## Contact

hello@kontinuity.space
https://kontinuity.space/privacy