The Complete Programmer's Guide to Coding Plans in 2026: China's Seven Major AI Coding Subscription Services Compared

Introduction: Understanding the Coding Plan Landscape

This comprehensive guide presents a complete comparison of China's seven major Coding Plan services (Volcengine, Alibaba, Tencent, Baidu, Kimi, Zhipu, and MiniMax), detailing each platform's supported models, pricing structures, quota allocations, and distinctive characteristics. This essentially provides a panoramic view of China's current AI coding ecosystem.

What Exactly Is a Coding Plan?

Coding Plans represent subscription packages for AI programming models offered by cloud providers.

Core Characteristics:

• Fixed monthly fees
• Substantial AI request quota allocations
• Compatible with Cursor, Claude Code, Cline, OpenClaw, and other development tools

In Essence: Coding Plan = AI programming model subscription service.

1. Volcengine Ark Coding Plan

Platform: Volcengine (ByteDance)

Subscription URL: https://volcengine.com/L/prz5yexYHB8/

Supported Models

• Doubao-Seed-Code
• DeepSeek-V3.2
• GLM-4.7
• Kimi-K2

Users can freely switch between models or utilize Auto mode for automatic selection.

Pricing Tiers

Distinctive Features

• Multi-model aggregation
• Auto model selection capability
• Compatible with Cursor / Claude Code / Cline
• Primarily focused on ByteDance ecosystem models

2. Alibaba Cloud Coding Plan

Platform: Alibaba Cloud

Subscription URL: https://bailian.console.aliyun.com/

Supported Models

• Qwen3.5-Plus
• Qwen3-Coder
• GLM-5
• MiniMax-M2.5
• Kimi-K2.5

Users can freely switch between multiple models.

Pricing Tiers

Distinctive Features

• Qwen programming models demonstrate exceptional strength
• Extensive model variety
• Robust Agent ecosystem
• Usage restrictions apply (AI Coding tool calls only)

3. Tencent Cloud Coding Plan

Platform: Tencent Cloud

Documentation: https://cloud.tencent.com/document/product/1772/128947

Supported Models

• Tencent HY 2.0
• GLM-5
• Kimi-K2.5
• MiniMax-M2.5

Pricing Tiers

Renewal Pricing:

• Lite: 40 CNY/month
• Pro: 200 CNY/month

Distinctive Features

Advantages:

• Tencent Hunyuan model integration
• Strong enterprise ecosystem

Disadvantages:

• Programming model maturity slightly weaker than Qwen

4. Baidu Qianfan Coding Plan

Platform: Baidu

URL: https://cloud.baidu.com/product/codingplan.html

Supported Models

• GLM-5
• MiniMax-M2.5
• Kimi-K2.5
• Wenxin series models

Distinctive Features

• Qianfan ecosystem integration
• Large enterprise user base
• Smaller developer community

5. Kimi Code

Platform: Moonshot AI

URL: https://www.kimi.com/code

Core Models

• Kimi K2
• Kimi K2.5

Distinctive Features

• Exceptional long-context capabilities
• Strong reasoning abilities

Ideal Use Cases

• Complex project analysis
• AI Agent development

6. Zhipu GLM Coding Plan

Platform: Zhipu AI

URL: https://www.bigmodel.cn/glm-coding

Supported Models

• GLM-4.7
• GLM-5

Distinctive Features

• Strong reasoning capabilities
• Robust Agent ecosystem

7. MiniMax Token Plan (Added 2026-03-26)

MiniMax has recently proven effective for programming tasks, serving as a perfect replacement for previously used ChatGPT workflows.

URL

https://platform.minimaxi.com/subscribe/token-plan

Supported Models

• MiniMax 2.7
• Speech-2.8-HD (Voice Model)
• MiniMax Hailuo 2.3 (Video Model)
• image-01 (Image Model)
• music-2.5+ (Music Model)

Distinctive Features

• M2.7 programming flagship model
• Exceptional cost-performance ratio
• Comprehensive multimodal one-stop solution

Pricing

Screenshot reflects monthly subscription pricing; annual packages offer additional savings. Minimum price reaches 29 CNY/month—note usage limits apply.

Comprehensive Six-Platform Coding Plan Comparison

Programming Capability Rankings

For programming capability benchmarks, refer to SWE-bench, where GLM-5, MiniMax-2.5, Kimi-2.5, and DeepSeek-3.2 have all achieved leaderboard positions.

Reference: https://www.swebench.com/

Personal Experience and Recommendations

Having consistently used ChatGPT for programming from version 5.0 through 5.3, I have now completely transitioned to domestic large models, primarily utilizing GLM and MiniMax. These have proven to be perfect ChatGPT replacements for my workflow.

Additional Note: DeepSeek currently lacks particularly outstanding differentiation; further updates will be provided as the landscape evolves.

Selection Framework

When choosing a Coding Plan, consider these factors:

1. Primary Use Case: Code generation vs. analysis vs. debugging
2. Budget Constraints: Monthly subscription vs. pay-per-use
3. Model Preferences: Specific model strengths for your workflow
4. Integration Requirements: Compatibility with existing tools
5. Quota Needs: Expected monthly usage volume

Conclusion

China's AI coding ecosystem has matured significantly, offering viable alternatives to international services. The competitive landscape benefits developers through diverse options at accessible price points.

Recommendation: Start with Lite tiers to evaluate platform fit before committing to Pro subscriptions. Many providers offer attractive first-month promotional pricing for trial purposes.

Note: Pricing and features subject to change. Verify current offerings through official channels. Corrections welcome for continuous updates.

Originally published on Zhihu: https://zhuanlan.zhihu.com/p/2015468530938693485