GLM Coding Plan

Zhipu AI's flat-fee coding subscription with 5-hour conversation windows and 20+ supported tools.

From¥49/mo Last verified:

Bottom line

GLM Coding Plan suits developers who want GLM-5 models in mainstream coding agents. Its core strength is the widest officially supported tool list in this category; the main constraint is the relatively tight entry-tier conversation allowance.

GLM Coding Plan is Zhipu AI's subscription for coding assistants on bigmodel.cn. For a flat monthly fee from ¥49, you get conversation-based quotas (roughly 80 chats per 5-hour window and ~400 chats per month at the entry tier) usable across an officially documented list of 20+ coding tools, including Claude Code.

Pricing and tiers

TierPriceModelsLimitsNotes
Lite¥49/mo3~80 prompts / 5h; ~400 prompts / weekEntry tier. Per official docs, monthly allowance ≈ 15–30× the subscription fee (at API prices, after the weekly cap).
ProNot officially specified3~400 prompts / 5h; ~2,000 prompts / weekMid tier. Price is listed on the official page (not published in the public docs we verified).
MaxNot officially specified3~1,600 prompts / 5h; ~8,000 prompts / weekTop tier. Price is listed on the official page (not published in the public docs we verified).

Quota and rate limits explained

Metering unit
Usage is metered by "prompts" (one prompt ≈ 15–20 model calls per official docs), within a 5-hour window plus a weekly window.
5-hour rolling window
A 5-hour window applies, recovering on a rolling basis. Lite ~80, Pro ~400, Max ~1,600 prompts per window.
Weekly quota
A weekly window (7-day cycle from purchase) also applies: Lite ~400, Pro ~2,000, Max ~8,000 prompts per week.
Monthly quota
Per official docs, the monthly allowance equals roughly 15–30× the subscription fee at API prices (after accounting for the weekly cap).
Concurrency
Not officially specified
Quota refresh
The 5-hour window refreshes 5 hours after consumption; the weekly window resets every 7 days.
When quota runs out
Per official docs: when the plan allowance is exhausted, you wait for the next 5-hour cycle — the system will NOT consume your other resource packs or account balance.
  • GLM-5.2 / GLM-5-Turbo are high-tier models: they consume quota at 3× during peak hours (14:00–18:00 UTC+8) and 2× off-peak. (Limited-time promo: 1× off-peak, until end of September.)
  • Prompt counts are estimates and vary with project complexity, codebase size, and whether auto-accept is enabled.

Supported models

GLM-5 series models are included; the official list covers 7 models at verification time.

Supported tools

  • Claude CodeOfficialOfficial setup guide is provided.
  • ClineOfficial
  • Roo CodeOfficial
  • OpenCodeCommunity
  • CursorCommunity
  • ContinueCommunity

Who it suits

Suitable for

  • Developers who want GLM-5 series models inside Claude Code or other coding agents
  • Users who value a wide officially supported tool list
  • Individual developers with moderate daily coding usage

Not suitable for

  • Heavy users who would exceed ~80 chats per 5-hour window
  • SaaS products that need server-side commercial API calls
  • Users who need non-GLM models under one subscription

Strengths and limitations

Advantages

  • Officially documents 20+ supported coding tools, one of the widest lists in this category
  • GLM-5 series models are competitive for agentic coding
  • Flat monthly fee with clear conversation-based quotas

Limitations

  • Entry-tier allowance (~80 chats / 5h) is relatively tight for heavy users
  • Only GLM first-party models are included
  • Overseas payment and signup requirements are not officially documented

Plan vs regular API

AspectGLM Coding PlanRegular pay-as-you-go API
BillingFlat monthly subscriptionPay-as-you-go per token
API key typeSubscription credential configured in coding toolsStandard platform API key
Supported tools20+ tools listed officially, incl. Claude CodeAny compatible client
Production useIntended for interactive codingSuitable for production workloads
Server-side callsNot the intended useSupported
ConcurrencyNot officially specifiedAccount-level rate limits apply
Quota rules5-hour window + monthly conversation allowanceNo window quotas; billed by actual usage
Best forDaily interactive coding sessionsAutomation, backends, batch jobs

Alternative plans

Related guides and reading

FAQ

How much does GLM Coding Plan cost?

The entry Lite tier starts at ¥49 per month. The official site lists 3 tiers (Lite / Pro / Max) with higher usage windows at higher prices.

What are the GLM Coding Plan usage limits?

At the entry tier, roughly 80 chats per 5-hour window and about 400 chats per month. The window recovers on a rolling basis and the monthly allowance resets each billing cycle.

Does GLM Coding Plan support Claude Code?

Yes. Zhipu AI provides official setup documentation for Claude Code, and lists 20+ supported coding tools in total.

Which models are included in GLM Coding Plan?

GLM first-party models, including the GLM-5 series (e.g. GLM-5.2, GLM-5.1). Third-party models are not included in the subscription.

Can the GLM Coding Plan credential be used as a normal API key?

The subscription is intended for interactive coding tools. For programmatic or production use, use the standard pay-as-you-go API on the bigmodel.cn platform.

Can overseas users buy GLM Coding Plan?

The plan is sold on the Chinese bigmodel.cn platform. Whether overseas phone numbers and international payment methods are supported is not officially documented — verify during registration.

What happens when I run out of quota?

The exact behavior after exhausting the window quota is not officially specified. In general you wait for the 5-hour window to recover, or upgrade to a higher tier for larger windows.

Notes
  • Prices and quota rules may change; this page was last verified on 2026-07-17.
  • The subscription is designed for interactive coding tools, not for production API workloads.

Sources and updates

Prices and quota rules may change at any time. This page was last verified on 2026-07-17. Always confirm on the official site before purchasing.

Sources

Update history

  • — Initial plan detail page created.
Last verified: