The plans at a glance
| Plan | Monthly price | How usage is metered | Models |
|---|---|---|---|
| GLM Coding Plan (Z.ai) | Lite $18, Pro $80, Max $168; 20% off quarterly, 30% off yearly | Credits per five hours and per week: Lite 2,000 and 10,000; Pro 6x and Max 14x Lite | GLM-5.3, GLM-5.3-Flash |
| Kimi Code (Moonshot) | Plus $19, Pro $39, Max $99, Ultra $199; less billed yearly | Rolling five-hour window plus a monthly total | K3.1 Pro and K3 on Plus and up |
| Alibaba Cloud Coding Plan | Pro $50 (Lite is closed to new subscribers) | Up to 6,000 requests per five hours, 45,000 a week, 90,000 a month | Qwen 3.7 Plus, Qwen 3.6 Plus, Kimi K2.5, GLM-5, MiniMax M2.5 |
| opencode Go | $10 ($40 Go Plus) | 20% of each model's monthly allowance per five hours, 50% per week | Kimi, GLM, MiniMax, Qwen, DeepSeek, GPT and Gemini models |
| Ollama Cloud | Pro $20, Max $100 | Token-metered credits: $60 a month on Pro, $300 on Max | Ollama's cloud models |
| Nexum Router | $5.50/week (about $22) | No hourly or weekly windows on Qwen, Xiaomi and Meta; $5 of DeepSeek V4.1 Flash a day | DeepSeek, Qwen, Xiaomi and Meta |
Three ways plans meter you
Before picking a plan, look at how your agent works: long sessions, large repositories and many tool calls each stress a different meter.
- Time windows: a share of usage per five hours plus a weekly cap. GLM's plan and opencode Go work this way, as Claude's own plans do. Long agent sessions hit the five-hour wall first
- Credits: a monthly pool spent at each model's token rate, as on Ollama Cloud. Big contexts drain it fastest, because agents resend the conversation every turn
- Request quotas: a count of requests per window, as on Alibaba's plan. Many small tool calls add up quickly
Which tools each plan works with
- GLM Coding Plan: ZCode, Claude Code, Codex, OpenCode, OpenClaw, Cline and more
- Alibaba Coding Plan: Claude Code, OpenCode, Cursor, Codex, Qwen Code, Cline, Kilo CLI and more
- Kimi Code: its own CLI, VS Code extension and desktop app; third-party tools count toward the same quota
- opencode Go: the opencode client
- Nexum Router: Claude Code natively, plus any OpenAI-compatible tool
What about MiniMax?
MiniMax has folded its coding plan into M Plan, a general subscription with Go, Explore and Build tiers covering text, image, audio and video models, with M3.1 Flash as the text model on every tier. Because usage is shared across those modalities, it does not compare directly with the coding plans above.
How to choose
- Lowest entry price per month: opencode Go at $10
- One model family, used heavily: GLM's or Kimi's own plan
- A broad open-model catalog in one subscription: Alibaba's plan or opencode Go
- An agent that runs all day without window resets: Nexum Router
Frequently asked questions
What is an AI coding plan?
A subscription that bundles model access for coding agents such as Claude Code and opencode, metered by time windows, credits or request quotas rather than billed per token.
How much does the GLM Coding Plan cost?
Lite is $18 a month, Pro $80 and Max $168, with 20% off quarterly and 30% off yearly billing.
Which coding plans work with Claude Code?
GLM Coding Plan, Alibaba's Coding Plan and Nexum Router all list Claude Code as supported.
Sources
Prices, limits and features were checked against these pages on 2026-10-29.