GLM9 min readReviewed Apr 20, 2026

GLM Coding Plan Pricing in 2026: Lite vs Pro vs Max

If you searched for GLM Lite vs Pro vs Max, start with this: GLM Coding Plan is a coding subscription, not a prepaid API balance. The decision is really about prompt limits, weekly caps, MCP quotas, supported tools, and whether you use GLM-5.1 enough to care about its higher usage weight.

Published Apr 19, 2026Updated Apr 20, 2026
  • As of April 20, 2026, GLM Coding Plan publishes approximate 5-hour prompt limits of 80, 400, and 1,600 for Lite, Pro, and Max.
  • The MCP monthly quotas published in the official docs are 100, 1,000, and 4,000 for Lite, Pro, and Max.
  • Coding Plan and general API billing are still separate routes with separate keys, limits, and expectations.
  • GLM-5.1 is supported across plans, but the default model path in tools does not always switch to it automatically.
Quick note: This guide is based on public docs and release pages, but you should still verify current pricing, limits, supported tools, and region-specific billing on the official source before you pay, subscribe, or integrate.

What you are really buying with GLM Coding Plan

If you are comparing GLM Lite, Pro, and Max, the first thing to fix in your head is what you are buying. GLM Coding Plan is a subscription for supported coding tools such as Claude Code, OpenClaw, OpenCode, and Cursor. It is not a prepaid API wallet.

That sounds like a small wording change, but it is why so many buyers get confused. Expect it to behave like general API credit and the pricing will feel misleading. Treat it like a tool-first coding plan and the rest of the page becomes much easier to read.

GLM DevPack route map infographic
For GLM, the core split is DevPack subscription, supported tools, and separate API billing. Source: Z.AI DevPack overview.

The pricing and limits that matter most

For most buyers, the important numbers are not every model name in the docs. The important numbers are the approximate 5-hour prompt limit, the weekly ceiling, and the MCP monthly quota.

GLM Coding Plan tiers as of April 20, 2026
PlanPublic price signalApprox. 5 hoursApprox. weekly limitMCP monthly quota
LiteFrom $18/month80 prompts400 prompts100
ProCheck current subscription page400 prompts2,000 prompts1,000
MaxCheck current subscription page1,600 prompts8,000 prompts4,000
GLM Coding Plan 5-hour prompt limits

These are the fastest way to estimate whether Lite, Pro, or Max fits your daily coding volume.

Liteabout 80

Best for light testing or low-volume use.

Proabout 400

The practical middle tier for many individual developers.

Maxabout 1,600

The heavy-use option for longer or more parallel workflows.

These are approximate prompt-based limits from the official DevPack overview, not exact token caps. Source: Z.AI DevPack overview.

BuyGLM shows package prices in USD. When a source page is published in CNY, the displayed value uses a fixed 1 USD = 8 CNY conversion and should still be checked against the live vendor page before payment.
Lite has a stable public starting price in the current docs. Pro and Max should be verified on the live subscription page because those are the numbers most likely to change.

Supported models vs default model behavior

The official DevPack docs state that the plans support GLM-5.1, GLM-5-Turbo, GLM-4.7, and GLM-4.5-Air. The trap is assuming that “supported” automatically means “selected by default in every tool.”

That is not how the public docs describe it. Some tools still default to GLM-4.7 or GLM-4.5-Air unless you explicitly move them to GLM-5.1. For a reader searching “why is GLM still using 4.7,” that is the answer.

The GLM model question most buyers really have
QuestionPublic answerWhat it means for a buyer
Which models does the plan support?GLM-5.1, GLM-5-Turbo, GLM-4.7, and GLM-4.5-AirYou do not need a separate tier just to unlock GLM-5.1
Will tools default to GLM-5.1 automatically?Not alwaysCheck the tool page or the GLM-5.1 guide before assuming
How does GLM-5.1 consume plan usage?The FAQ still documents higher usage multipliers for 5.1 / 5-TurboHeavy GLM-5.1 use can feel more expensive in practice than the plan name suggests

Which tools are officially supported and what buyers get wrong

GLM’s tool story is broad. The official docs cover Claude Code, OpenClaw, OpenCode, Cursor, Coding Tool Helper, and more. That is one of the strongest reasons to buy into the route.

  • Do not treat Coding Plan usage as general API credit.
  • Do not assume every tool defaults to GLM-5.1 just because the plan supports it.
  • Do not ignore MCP quotas if your workflow uses search, reading, or vision heavily.
  • Do not rely on screenshots when the current tool page or FAQ answers the same question directly.

Use the plan page for the price, the overview for the limits, and the tool page for setup

That is the cleanest way to keep a GLM buying article accurate and genuinely useful to readers.

Sources and official links

Frequently asked questions

Does GLM Coding Plan include general API usage?

No. The official FAQ still treats Coding Plan and general API billing as separate routes.

How should I choose between Lite, Pro, and Max?

Choose by your likely 5-hour volume, weekly ceiling, and MCP usage. The tier names are less useful than those three numbers.

Why do some GLM tools still seem to use GLM-4.7 by default?

Because support and default selection are different. The public docs still show cases where the default path is not GLM-5.1 until you change it.

Where should I verify the final price before buying?

On the live GLM subscription page. That is the final checkpoint for anything that can change more often than the overview docs.