Practical guides for people who build with AI models and APIs
- 492 articles
- 6 languages
- 12 topics
- Updated Oct 10, 2026
What we cover
From choosing a model and estimating the bill to debugging a 429 in production, these are the topics we write about.
- AI Image Generation110Image models and APIs compared, with prompts, pricing, and editing workflows.
- API Guides63Integrating AI APIs: authentication, endpoints, SDKs, and production patterns.
- Claude Code54Setup, pricing, limits, workflows, and fixes for Anthropic’s Claude Code agent.
- Developer Tools & Agents50Coding agents, IDE assistants, local models, and AI developer workflows.
- Model Comparisons42Side-by-side comparisons of AI models for real tasks, costs, and limits.
- AI Video Generation33Video models such as Sora, Veo, and Kling: access, pricing, and quality trade-offs.
- Troubleshooting32Error codes, rate limits, and access problems with verified fixes.
- API Pricing & Plans31Token prices, subscriptions, quotas, and cost calculations for AI APIs.
- ChatGPT & OpenAI30ChatGPT plans, OpenAI models, Codex, and OpenAI API guides.
- Claude & Anthropic29Claude models, plans, API access, and Anthropic platform guides.
- Gemini & Google AI15Gemini models, Google AI Studio, Vertex AI, and Gemini API guides.
- AI Tools & Guides3Everyday AI tools, free options, and how-to guides.
How we write
Every guide follows the four rules below, so you can check our work instead of taking it on trust.
Start from a real task
Each guide answers one concrete question: which model to use, what next month's bill will look like, how to fix a specific error, or how to ship an integration.
Sources sit next to claims
Links to official docs and primary sources appear right after the statement they support, not in a list at the end. You can check a single number without rereading the whole guide.
Dates on anything that changes
Prices, quotas, rate limits, and availability go stale fast, so they carry an explicit "as of" date. Every page shows when it was published and, if revised, when it was last updated.
We say whether we tested it
We state whether a conclusion comes from official documentation or our own testing, and say so plainly when we did not run paid tests. Recommendations and inferences are labeled as such.
Who runs this blog
LaoZhang AI Blog is a third-party guide to AI models and tools: how to use them, what they cost, how to fix their errors, and what changed. We are not affiliated with OpenAI, Anthropic, Google, or other model vendors. The same team also runs laozhang.ai, a multi-model API gateway. Wherever a guide mentions laozhang.ai, we say it is our own product, its prices are platform pricing rather than an independent comparison, and the promo card is labeled as ours.
About bylines
Most guides are signed "LaoZhang AI Team", which means the piece went through our shared research, writing, and review process. The byline names who is responsible for the content.
LaoZhang AI TeamSpotted a mistake?
Email hi@laozhang.ai with the page URL, the exact sentence that is wrong, and a source we can use to verify the correction.
We check the evidence before changing anything. Significant corrections show up in the update date or as a note in the article; sending feedback doesn't guarantee a change.
For general questions, find us on Telegram:@laozhang_cn