The 30-second summary
+ What we liked
- Code-focused relay with good Claude support
- Competitive pricing
- Active updates
− What we didn't
- Newer station with limited history
- Focused on coding use cases only
In-depth review
Works without VPN from mainland China. Pays via 支付宝 or 微信支付 — no foreign card needed.
aicodemirror isn’t trying to be the biggest relay station. It’s targeting one specific user: the developer who wants to run Claude Code without a VPN and doesn’t want to pay OpenAI-level markup for GPT-4o. Two models, zero fluff.
Models & Tokens
You get exactly two models:
- GPT-4o
- Claude 3.5 Sonnet
That’s it. No Gemini, no DeepSeek, no experimental previews. The 100,000 max token context window is generous enough for multi-file code analysis or long conversation histories in Claude Code. If you need vision or multimodal, you’ll have to look elsewhere — both models support vision but the relay doesn’t advertise it as a focus.
The name “aicodemirror” isn’t a joke. This is a coding-only relay. Don’t come here for DALL-E, video generation, or general chat.
Pricing & Payment
| Plan | Price | Notes |
|---|---|---|
| Free trial | $0 | No credit card needed to start |
| Pay-as-you-go | Variable | Priced per model, competitive vs. direct API |
Pricing details aren’t published on the site upfront — you register and see rates per token. From community reports, GPT-4o and Claude 3.5 Sonnet rates are noticeably cheaper than OpenRouter or direct API, especially when paying in CNY. No minimum recharge amount specified, which is flexible but uncertain.
Payment is the selling point for mainland devs: 支付宝 and 微信支付 both work. No foreign credit card required. No USDT conversion. You top up in CNY and use the relay.
Performance & Reliability
Uptime sits at 94.0%. That’s not great. For a coding tool, a 6% downtime window can kill a debugging session. The safety rating is 4/5 — no content filtering horror stories yet, but the uptime is the real concern.
Latency from Beijing/Shanghai is acceptable for code completion and conversational debugging. Expect ~800ms–1.5s for first token on Claude 3.5 Sonnet, which is fine for non-real-time use. Don’t use this for latency-sensitive streaming chat.
Pros & Cons
Pros
- Code-focused relay with good Claude support
- Competitive pricing vs. direct API or OpenRouter
- Active updates — the team pushes changes regularly
- 支付宝 and 微信支付 support
Cons
- Newer station with limited history and community trust
- Only two models — no fallback for non-coding tasks
- Uptime at 94% — expect occasional outages
- No refund policy specified
Verdict
aicodemirror solves a real problem: running Claude Code from mainland China without VPN. The 支付宝/微信支付 support makes onboarding instant. But the limited model selection and 94% uptime mean this isn’t a primary relay — it’s a secondary option for when you need Claude without the hassle.
If you’re a developer who exclusively uses GPT-4o and Claude 3.5 Sonnet for coding, this works. If you need model diversity or high reliability, keep looking.
FAQ
Q: Does aicodemirror work without VPN from mainland China? A: Yes. No VPN required. Access is direct from mainland networks.
Q: Can I pay with WeChat Pay or Alipay? A: Yes. Both 微信支付 and 支付宝 are supported. No foreign credit card needed.
Q: How many models are available? A: Two: GPT-4o and Claude 3.5 Sonnet. No other models.
Q: What is the uptime guarantee? A: 94.0% uptime based on available data. No SLA published.
Q: Is there a free trial? A: Yes. Registration is free with $0 upfront. You pay per token after that.
Pricing breakdown
aicodemirror offers competitive pricing for developers. Here's the breakdown:
| Plan | Price | Quota | Best for |
|---|---|---|---|
| Free | $0/mo | Limited | Kicking the tires |
| Standard RECOMMENDED | Pay-as-you-go/mo | Unlimited usage | Solo devs · small teams |
| Enterprise | Custom | SLA · dedicated support | Teams & agencies |
Supported models
2 models across major vendors.
Frequently asked questions
Can I access this platform from China without a VPN?
Most relay stations are accessible from Chinese ISPs. Check our review for specific routing details.
What payment methods are accepted?
Payment options vary by platform. Some accept Alipay/WeChat Pay, others are USD/crypto only.
How does this compare to using OpenAI directly?
Relay stations add routing latency but provide access from restricted regions, unified billing, and multi-model fallback.
Is my API key safe?
Keys are encrypted at rest. Most platforms support per-project scoping and IP allow-lists.
Should you use aicodemirror?
Claude Code users wanting variety