My current AI strategy: Claude for heavy coding, ChatGPT for general-purpose tasks and voice, Gemini for multimodal and supplementary needs, GitHub Copilot for in-editor completions. If you're on a budget, one domestic model as a fallback is enough.
Claude
Best for terminal-based coding, complex refactors, and long-context collaboration.
View Usage GuideChatGPT
Best for voice, daily Q&A, cross-platform use, and as a general-purpose backup.
See My ChoicesCopilot
Best for high-frequency completions, chat, and Agent workflows inside VS Code.
View Copilot GuideDomestic Models
When budget is tight, Doubao or DeepSeek can serve as a lightweight backup.
See ComparisonTL;DR
If you're a developer
- Primary: Claude
- In-editor completions: GitHub Copilot
- Backup: ChatGPT or Gemini
If you're a general user
- Primary: ChatGPT
- Multimodal supplement: Gemini
- Budget option: Domestic models as fallback
AI Tools I Currently Use
Claude
Terminal coding, complex refactoring, long-context collaboration
If you regularly make real code changes in local repos rather than just chatting, Claude is still my top recommendation. Its strength isn't just "can it answer" — it's "can it keep pushing tasks forward within a project context."
ChatGPT
Daily Q&A, voice, multi-platform experience, general-purpose backup
If you're not a heavy code user, or you value voice, mobile, desktop, and general capabilities, ChatGPT is still the most reliable general-purpose choice. I use it myself as a backup to Claude.
Gemini
Multimodal, frontend UI tasks, third choice alongside Claude/ChatGPT
In my usage, Gemini is more of a supplementary tool rather than a sole primary. But it's valuable in multimodal tasks, certain frontend scenarios, and model-switching strategies. If you already use Claude or ChatGPT, Gemini is often a good third tool.
GitHub Copilot
In-editor high-frequency completions, chat, Agent, MCP
Copilot's value is being "closest to the code." If your main workflow lives in VS Code, it lets you complete, ask questions, and handle Agent scenarios with lower context-switching cost. I still keep it long-term.
Domestic Model Fallbacks
Doubao, DeepSeek etc. — focused on low cost and availability
If you're not ready to subscribe to multiple overseas AI services, domestic models can fill the gap. They're not my top primary recommendation, but they have value for translation, lightweight Q&A, and budget control.
How to Choose by Use Case
Just pick one
General users: ChatGPT first. Developers: Claude first.
Want both coding and daily use covered
`Claude + ChatGPT` is my recommended dual-subscription combo.
Mainly working inside VS Code
Start with Copilot, then decide if you need Claude or ChatGPT on top.
On a tight budget
Pick one primary, add a domestic model as fallback. Don't subscribe to everything at once.
Related Guides
AI Services Comparison
Start here to quickly understand which services I actually use.
Claude Code Guide
For users ready to start coding with the CLI.
GitHub Copilot in VS Code Guide
For VS Code users looking to build an AI coding workflow quickly.
Claude Pro Low-Price Region Subscription
Read this after you've decided to subscribe to Claude.
Claude Pro US Subscription
A direct record of the Claude subscription process.
Making ChatGPT Your iPhone Voice Assistant
For users interested in mobile voice and productivity workflows.
One-Line Advice
Don't subscribe to every AI tool right away. Start with one primary based on your main use case, then decide if you need a second or third — this keeps costs low and the experience clear.
