alternative · Z.ai GLM Coding Plan
A Z.ai GLM Coding Plan alternative built for private coding agents
Z.ai GLM Coding Plan is a solid option for many uses. If you want the best open coding models behind one zero-retention endpoint that Claude Code and other agents can use as-is, Hushwire is built for exactly that. Hushwire serves GLM alongside Qwen, Kimi and DeepSeek through one zero-retention endpoint with a written routing policy you can show a client.
Side by side
| Z.ai GLM Coding Plan | Hushwire | |
|---|---|---|
| What it is | A cheap, Anthropic-compatible coding plan for GLM models from the GLM lab itself. | A private, drop-in backend for coding agents. |
| Retention | Governed by the lab’s own terms; many client contracts rule out sending code to China-based providers. | Zero retention on every request, no toggle. Metadata only, 30 days, visible to you. |
| Pricing | Low flat monthly plans. | Flat: $19 / $49 / $149 a month with a fair-use allowance; $5 per 10M top-up. |
| Claude Code | Varies. | Native Anthropic /v1/messages with tool use and streaming. |
| Big-lab routing | Varies. | Never routes to OpenAI or Anthropic models; open-weight only. |
| Client paperwork | — | Downloadable security summary and data-flow diagram. |
Summaries of public information as of November 2026. Check Z.ai GLM Coding Plan’s current terms before deciding.
Switching takes two environment variables
Claude Code
# Powered by Hushwire private inference (zero retention): hushwire.vibiz.app
export ANTHROPIC_BASE_URL="https://hushwire.vibiz.app"
export ANTHROPIC_AUTH_TOKEN="hw_your_key_here"
export ANTHROPIC_MODEL="hush-coder-large"
export ANTHROPIC_DEFAULT_OPUS_MODEL="hush-coder-large"
export ANTHROPIC_DEFAULT_SONNET_MODEL="hush-coder-large"
export ANTHROPIC_DEFAULT_HAIKU_MODEL="hush-coder"
export CLAUDE_CODE_SUBAGENT_MODEL="hush-coder-large"
export CLAUDE_CODE_DISABLE_NONESSENTIAL_TRAFFIC=1
claudeUsing Cline, OpenCode or Aider? See every agent’s config.
Models
| Alias | Model | Context | Good for |
|---|---|---|---|
| hush-coder | Qwen3 Coder 30B A3B (Qwen) | 262k | Fast, cheap default for edits, refactors and tool loops. |
| hush-coder-large | Qwen3 Coder Next (Qwen) | 262k | The largest Qwen coder. Best for multi-file agent work. |
| hush-kimi | Kimi K2.5 (Moonshot AI) | 262k | Strong agentic coder with long tool chains. |
| hush-glm | GLM 4.7 (Z.ai) | 200k | Reliable tool calling, good at planning and front-end work. |
| hush-deepseek | DeepSeek V4 Flash (DeepSeek) | 128k | Very cheap general coder, good for quick questions. |
Try it before you configure anything
The playground runs the same open coding models. No sign-up, 20 messages a day, and nothing you type is stored.