About LLMScout.tech
LLMScout.tech is a free set of calculators for developers and founders who need a quick, honest answer to "how much will this AI feature actually cost?" before they build it. Every tool runs in your browser, requires no account, and links back to official provider documentation so you can verify every number we show.
Why this exists
AI API pricing is scattered across provider docs, changes often, and is easy to get wrong when you're estimating a new feature. We built LLMScout.tech to put OpenAI, Anthropic, Gemini, GLM and Kimi pricing side by side, backed by live pricing data where possible, so you can plan a budget in minutes instead of digging through multiple pricing pages. The site grew out of a practical frustration: budgeting an LLM feature used to mean juggling per-token rates, tokenizer quirks and currency conversion in an improvised spreadsheet, where small mistakes compounded into badly wrong estimates. These calculators encode that spreadsheet once, correctly, and keep the prices behind it fresh.
Who maintains it
LLMScout.tech is built and maintained by LW Forge, a software studio that works on web products and LLM API integrations — OpenAI, Anthropic and others — in real applications. The calculators began as internal spreadsheets for estimating API budgets on the studio's own projects, and became public tools once it was clear the same cost questions kept coming up everywhere. LW Forge maintains the site's code, pricing data pipeline and content directly, and reviews every change that affects a displayed price.
Who writes and reviews the content
Luciano Cordeiro
Software engineer · creator of LLMScout.tech · LW Forge
Luciano Cordeiro is a senior software engineer based in Brazil with more than five years of experience building backend services — APIs and microservices — and front ends in React and Vue.js, working mainly with JavaScript, Node.js, Python and FastAPI. He created LLMScout.tech at LW Forge to replace the spreadsheets he used to estimate LLM API budgets. On the blog he writes about OpenAI's models, tokenization, what the APIs cost in Brazil and the mechanics behind the bill — caching, batch discounts, reasoning tokens. He decides what the site publishes and answers for corrections.
Wallace Gomes de Souza
Software engineer · LW Forge
Wallace Gomes de Souza is a software engineer at LW Forge. He built the publishing pipeline behind the LLMScout.tech blog — the workflow that prepares each post, runs the automated content audit and hands it over for review. On the blog he writes about Anthropic's Claude family, open-weight models such as GLM, Kimi and Nemotron, Google's Gemini line, GPU infrastructure and context windows; his coverage of Claude Sonnet 5's permanent pricing included the correction of nine earlier articles that still quoted the old rate.
Pricing methodology
Model prices are fetched from OpenRouter's public pricing API, which aggregates the official per-token rates published by OpenAI, Anthropic, Google, Z.ai, Moonshot AI, NVIDIA and other providers. We cache that feed for 24 hours, so under normal operation every price you see is at most one day old. USD/BRL exchange rates come from AwesomeAPI on a 1-hour cache, refreshed in the background while you use the site.
If the live feed is unavailable, calculators fall back to a hardcoded price table that we review manually against official provider pricing pages. Fallback values are conservative snapshots, not live data — one more reason every calculator links to the provider's own documentation. Token counts are estimates based on typical tokenizer behavior (roughly 4 characters or 0.75 words per token in English); real counts vary by model and language, so treat results as planning figures, not invoices.
Found a price that looks wrong or out of date? Pricing corrections are the highest-priority fix on this site. Email contato@lwforge.com with the calculator, the model and the value you expected, and we'll check it against the provider's official page — usually within a couple of days.
Independence
LLMScout.tech is an independent project and is not affiliated with, endorsed by, or sponsored by OpenAI, Anthropic, Google, Z.ai, Moonshot AI or NVIDIA. All trademarks belong to their respective owners.
Contact
Questions, corrections or feedback: contato@lwforge.com.