Description
GLM AI is Zhipu AI’s family of large language models, developed under the international brand Z.ai and led by the GLM-5.2 flagship model. The chat interface at chat.z.ai is free to use with no subscription or per-token billing, giving general users direct access to GLM’s reasoning and multimodal capabilities. Developers can also access GLM models through a pay-per-token API, with GLM-5.2 priced well below comparable Western frontier models, and several smaller models, including GLM-4.7-Flash and GLM-4.5-Flash, offered free on the API. For coding-tool users, Z.ai sells a separate GLM Coding Plan, a flat monthly subscription (Lite, Pro, and Max tiers) that gives a fixed prompt quota inside supported IDEs and CLI tools like Claude Code, Cline, and Roo Code, rather than metering every token. Several GLM models are released as open weights, letting organizations self-host on their own hardware instead of relying on Zhipu’s API. It suits everyday users who want a free, capable chat assistant, and developers who want a lower-cost, open alternative to Western frontier models for coding and agentic workloads.
Key Features
- Free chat interface — chat.z.ai gives full access to GLM models with no subscription or per-token billing.
- GLM-5.2 flagship model — frontier-class reasoning and agentic model, priced well below comparable Western models on the API.
- Free API models — GLM-4.7-Flash, GLM-4.5-Flash, and GLM-4.6V-Flash are offered at $0 for input, output, and cached input.
- GLM Coding Plan — flat monthly subscription for using GLM models inside supported coding tools, billed by prompt quota rather than tokens.
- Open-weight releases — several GLM models are released publicly, allowing self-hosting instead of relying on Zhipu’s own API.
- Long context window — GLM-5.2 supports up to a 1 million-token context window on the API.
- Cached input pricing — repeated context is billed at a steep discount compared to standard input tokens.
How It Works
Everyday users interact with GLM models for free through the chat.z.ai web interface, similar to using a free ChatGPT or Claude account, with no subscription or credit card required. Developers building their own applications instead use the Z.ai API, billed per million tokens by model, with GLM-5.2 as the flagship reasoning option and several smaller Flash models offered entirely free for lighter workloads. For developers who work inside coding tools like Claude Code, Cline, or Roo Code rather than calling the API directly, Z.ai sells the GLM Coding Plan, a flat monthly subscription that provides a prompt quota refreshing on a 5-hour and weekly cycle instead of metering individual tokens. The Coding Plan only works inside officially supported tools, not as a general-purpose API replacement. Organizations with steady, high-volume usage can also self-host several GLM models using their own hardware, since Zhipu publishes the weights for select releases rather than keeping them entirely proprietary.
Technical Architecture & Overview
- Core Engine: GLM-5.2, Zhipu AI’s flagship large language model, alongside GLM-5-Turbo, GLM-4.7, and lightweight Flash variants; several GLM releases ship as open weights under permissive licenses.
- Deployment: Free web chat at chat.z.ai, a pay-per-token developer API, integration inside 20+ supported coding tools through the GLM Coding Plan, and self-hosting for organizations using open-weight releases.
- API Surface: GLM-5.2 API priced at $1.40 per million input tokens and $4.40 per million output tokens, with cached input billed separately at a steep discount; GLM-4.7-Flash and GLM-4.5-Flash are free on the API.
- Known Limits: The GLM Coding Plan works only inside officially supported coding tools, not as a general API key replacement; GLM-5.2 and GLM-5-Turbo consume prompt quota at 3x during peak hours and 2x off-peak on the Coding Plan.
Pros & Cons
| Pros | Cons |
|---|---|
| chat.z.ai is genuinely free with no subscription, credit card, or per-token billing required | The GLM Coding Plan is restricted to officially supported coding tools and cannot replace a general Z.ai API key |
| GLM-5.2 API pricing runs well below comparable Western frontier models on a per-token basis | Pro and Max GLM Coding Plan prices have shifted with promotions since launch, so current rates should be confirmed directly before subscribing |
| Several models (GLM-4.7-Flash, GLM-4.5-Flash, GLM-4.6V-Flash) are free on the API, not just trial credits | GLM-5.2 and GLM-5-Turbo consume coding-plan quota at up to 3x the rate of older models during peak hours |
| Multiple GLM releases ship as open weights, allowing self-hosting instead of relying on Zhipu’s own infrastructure | Enterprise pricing is not published; organizations must contact Z.ai sales directly |
| GLM Coding Plan’s Max tier is priced below both ChatGPT Pro and Claude Max’s top tier for comparable heavy agentic coding use | As a China-based provider, some regulated or latency-sensitive US-based workloads may be better served by a domestic alternative |
Pricing
| Plan | Price |
|---|---|
| Chat (chat.z.ai) | Free, no subscription required |
| API (GLM-5.2 and other models) | Token-based pricing |
| GLM Coding Plan (Lite) | $18/month (introductory pricing may apply) |
Platform Availability
Web (chat.z.ai) | API | Coding tool integrations (Claude Code, Cline, Roo Code, and 20+ others)
Best For
Everyday free chat users | Budget-conscious developers | Agentic coding workflows
Frequently Asked Questions
Is GLM AI free to use?
Yes, for chat. chat.z.ai is free to use with no subscription or per-token billing, according to Z.ai’s own product information; the API and GLM Coding Plan are separate, paid products for developers.
Does GLM AI have any free API models?
Yes. GLM-4.7-Flash, GLM-4.5-Flash, and the vision model GLM-4.6V-Flash are priced at $0 for input, output, and cached input on Z.ai’s official API pricing page.
What is the GLM Coding Plan?
It’s a flat monthly subscription (Lite, Pro, and Max tiers) that gives a fixed prompt quota for using GLM models inside supported coding tools like Claude Code and Cline, rather than billing per token, according to Z.ai’s official documentation.
Are GLM models open source?
Several are. Zhipu AI releases select GLM models, including lightweight Flash variants, as open weights under permissive licenses, allowing self-hosting instead of relying on the Z.ai API.
How does GLM AI differ from Kimi AI?
Both are Chinese AI labs offering a free chat interface alongside a separate paid coding subscription, but GLM’s chat access at chat.z.ai is entirely free with no tier structure, while Kimi’s free Adagio tier sits below four paid membership tiers with scaling agent and context limits.
Similar Tools in This Directory
Also in this directory: ChatGPT, OpenAI’s assistant with bundled Codex access | Claude by Anthropic, coding agent bundled into every paid plan | Google Gemini AI, multimodal assistant integrated into Google Workspace | Meta AI, free assistant built into Facebook, Instagram, WhatsApp | Grok by xAI, real-time X and web search built into chat | Perplexity AI, cited-answer research engine | Kimi AI, Moonshot’s agentic workspace with a free tier
