Skip to content

Latest commit

 

History

History
66 lines (53 loc) · 3.46 KB

File metadata and controls

66 lines (53 loc) · 3.46 KB

GLM Coding Plan — Main LLM 🧠

Pricing

  • Lite: $3/month (first month/quarter/year paid upfront), then $6/month
  • Pro: $15/month (first month/quarter/year paid upfront), then $30/month
  • Max: $30/month (first month/quarter/year paid upfront), then $60/month
  • Practicality of limits: Very hard to hit even with 2–5 parallel agents on Pro and Max plans
  • Lite: 120 prompts per 5 hours
  • Pro: 600 prompts per 5 hours
  • Max: 2,400 prompts per 5 hours
  • Note: I'm on Max plan currently, but haven't hit limits when I was on Pro plan for over a month in the past aswell
  • IMPORTANT Integration Tools: Claude Code and Droid CLI are generally on par when it comes to efficiency in most cases. Main differences:
    • Claude Code: Faster to work with and iterate on code from my experience
    • Droid CLI: Better planning mode overall for complex architectures
    • Tip: Planning capabilities can be improved for both tools using Clavix for prompt optimization
  • Optional: 10% off referral link (disclosure): https://z.ai/subscribe?ic=CUEFJ9ALMX

Why GLM?

  • Lowest cost while keeping quality around modern, commercial LLMs level
  • State‑of‑the‑art open‑source model
  • Can autonomously resolve bugs for 10–15 minutes without intervention
  • Handles complex projects (Next.js + complex database + backend)
  • Works quite well with modern web development stacks (Svelte, React, Next.js, Astro, D1 databases, Cloudflare Wrangler files, etc.)

Cost comparison

  • Claude Code Max20: €200+ per month (with EU VAT)
  • Codex Pro: €229 in EU with VAT per month
  • Most alternatives will be more expensive than GLM considering the fact that GLM Coding Plan efficiently gives you A TON of tokens per 5h, not even talking about per month.

Recommended Workflow Integration

Best Used In:

Budget Considerations:

Integration Setup:

Workflow Phases by Usage:

Phase 0 (Preparation):

Phase 1 (Planning):

  • Use for PRD refinement and feature breakdown
  • Architecture discussions for Core Technologies stack

Phase 2 (Development):

  • Daily coding assistance and debugging
  • DevTools MCP integration for automated testing
  • Component generation for Astro + Tailwind projects

Phase 3 (Testing & Debugging):

  • Bug analysis and resolution
  • Code review and optimization
  • Integration with testing workflows

Back: AI Model Providers