GLM Coding Plan Provider for Claude Code

GLM-5.2 is a frontier-class reasoning model accessible through z.ai’s Coding Plan. It exposes a native Anthropic-compatible endpoint, so Claude Code talks to it without adapters or middleware.

Model specs

Thinking mode is enabled by default. The template sets REASONING_EFFORT to high and allocates 8,000 thinking tokens, which is enough for most code tasks without burning through your context window.

Context windows are mixed across tiers, so the template exposes them per-model rather than with a single global override. GLM-5.2 carries a [1m] suffix that tells Claude Code its real 1M window, while Claude Code’s default 200K assumption for unrecognized models already matches GLM-5.1 and GLM-5-Turbo exactly — so auto-compaction triggers at the right point for every tier without any CLAUDE_CODE_AUTO_COMPACT_WINDOW override.

Setup

1. Get a Coding Plan subscription at z.ai
2. Copy your API key from the dashboard
3. Run the setup command:

claude-multi add glm

4. Paste your API key when prompted

That is the whole process. The template configures the base URL, model mappings, context limits, and thinking parameters. Your instance is ready immediately.

When to pick GLM

GLM-5.2 is a good fit when you want a fixed monthly cost instead of per-token billing. The Coding Plan gives you a generous allocation of requests, and GLM-5-Turbo handles lighter tasks (quick edits, shell commands, subagent work) at higher speed, while GLM-5.1 covers everyday coding in between.

If your workload is bursty and you prefer paying only for what you use, look at the DeepSeek or MiMo pay-per-token templates instead.

Pricing details

GLM uses a Coding Plan subscription model. You pay a flat monthly fee and get an allocation of requests. Check z.ai for current pricing tiers.

Related providers

• DeepSeek - pay-per-token, also frontier coding
• MiniMax - 1M context window, subscription
• Qwen - Alibaba’s coder models, pay-per-token