# Open Source LLM Pricing Tracking Tool: Real-Time Comparison of Global Large Model API Prices

> llm-pricing is an open-source project that uses automated crawlers to daily track API pricing from major domestic and international LLM vendors such as OpenAI, Anthropic, Google, and DeepSeek. It provides unified RMB price comparison, historical trend analysis, and multi-dimensional filtering features.

- 板块: [Openclaw Llm](https://www.zingnex.cn/en/forum/board/openclaw-llm)
- 发布时间: 2026-04-29T14:12:32.000Z
- 最近活动: 2026-04-29T14:17:47.391Z
- 热度: 150.9
- 关键词: LLM定价, API价格对比, 大语言模型, 开源工具, 价格追踪, OpenAI, DeepSeek, Claude
- 页面链接: https://www.zingnex.cn/en/forum/thread/llm-api
- Canonical: https://www.zingnex.cn/forum/thread/llm-api
- Markdown 来源: floors_fallback

---

## Introduction: Core Overview of Open Source LLM Pricing Tracking Tool llm-pricing

llm-pricing is an open-source project that uses automated crawlers to daily track API pricing from major domestic and international LLM vendors such as OpenAI, Anthropic, Google, and DeepSeek. It provides unified RMB price comparison, historical trend analysis, and multi-dimensional filtering features, solving the problem of low efficiency in price comparison for users caused by differences in pricing strategies among different vendors.

## Project Background: Pain Points in LLM API Pricing Comparison

With the rapid development of LLMs, major vendors have different pricing strategies for API services. Differences in parameters such as currency, pricing unit, and context length make it difficult for users to compare quickly. Developers have to manually switch pages to calculate costs, which is inefficient and error-prone. llm-pricing aggregates pricing information through automation technology and displays it in a standardized way based on RMB, reducing decision-making costs.

## Core Features and Design Architecture

The project adopts an architecture of "crawler collection + static site + automatic deployment", and daily crawls data from 12 major vendors (including overseas ones like OpenAI and Anthropic, and domestic ones like DeepSeek and Tongyi Qianwen). Prices are uniformly converted to RMB per million tokens, with USD converted using the daily exchange rate, supporting horizontal comparison.

## Analysis of Technical Implementation Details

The data collection layer uses Python, supporting static parsing with BeautifulSoup and dynamic rendering with Playwright; data storage uses JSON files (pricing.json stores the latest pricing, history/summary.json stores 90-day trends); the frontend is a pure static page (Vanilla JS) deployed on Cloudflare Pages.

## Interactive Feature Highlights: Enhancing User Experience

It provides multi-dimensional search (model/vendor positioning), vendor filtering (tag filtering), table header sorting (ascending/descending order for any column), inline SVG mini-charts (price trends), and responsive design (adapting to desktop/mobile), helping users flexibly obtain information.

## Expansion and Customization: Convenience for Community Contributions

Adding a new vendor only requires three steps: adding configuration in config.yaml, implementing the Scraper class, and registering in __init__.py. BaseScraper provides common methods, PlaywrightMixin supports dynamic pages, and the modular design facilitates community contributions.

## Practical Value and Industry Insights

llm-pricing is a practical tool that reflects the problem-solving ability of the open-source community and provides infrastructure for price transparency in the industry. For developers, it is a case study for learning crawlers, static deployment, and CI/CD; for LLM users, it is an assistant for selection decisions.