The 30-second summary
+ What we liked
- API aggregation focus
- Low barrier to entry
− What we didn't
- Limited independent verification data
- Small community footprint
In-depth review
Works without VPN from mainland China. Accepts Alipay and WeChat Pay directly — no overseas credit card needed.
接口AI positions itself as a low-friction relay for Chinese developers who want GPT-4o, Claude 3.5, or Gemini 2.0 without the VPN dance. The platform is currently in a trial phase (free, no pricing tier), which makes it a zero-cost test for anyone curious about API aggregation. But free comes with caveats.
Model Support
The model list is deliberately small: GPT-4o, Claude 3.5, and Gemini 2.0. No Mistral, no DeepSeek, no Llama variants. That’s it. If you need anything outside those three, you’re out of luck.
- GPT-4o — OpenAI’s current flagship multimodal model. Works as expected.
- Claude 3.5 — Sonnet tier. Good for long-context reasoning and code generation.
- Gemini 2.0 — Google’s latest. Latency is reasonable from Beijing.
No model version suffixes (e.g., “GPT-4o-2024-08-06”) are visible in the platform data. If you need precise model pinning, check before building on top of this relay.
Pricing
| Plan | Price | Notes |
|---|---|---|
| Free trial | $0 | No time limit mentioned in data |
| Paid tiers | Not announced | No pricing info available |
There is no recharge minimum, no refund policy, and no promo code. The platform is effectively in open beta with zero cost. That’s rare for a Chinese relay. Most competitors charge ¥10–¥50 minimum recharge. 接口AI currently asks for nothing.
Payment Methods
Alipay and WeChat Pay are the only options. No crypto, no international cards. That’s standard for a China-facing relay — and a relief if you’re paying in CNY.
Uptime & Reliability
Uptime is listed at 94%, which is low compared to established relays like API2D (99%+) or AiHubMix (98%+). A 6% downtime window means you’ll see errors roughly 1–2 days per month. For a free trial, that’s acceptable. For production use, that’s a red flag.
Safety rating: 3/5. The platform doesn’t publish independent uptime monitoring or SLA guarantees. Composite score is 56.4/100 — below average for the category.
Latency from China
Based on the platform data alone, no specific latency numbers are provided. But given the model set (GPT-4o, Claude 3.5, Gemini 2.0), typical relay latency from Beijing/Shanghai to these models runs 800–1500ms per request. 接口AI likely routes through a domestic proxy, so expect higher latency than direct API calls but lower than a VPN + direct call.
Pros & Cons
Pros
- Zero cost to start — no recharge, no subscription
- Works without VPN from mainland China
- Alipay and WeChat Pay accepted
- API aggregation reduces the need for multiple API keys
Cons
- Only 3 models available
- 94% uptime is below industry average
- No refund policy or SLA
- Small community footprint — harder to find support or documentation
- No pricing transparency for future paid tiers
Verdict
接口AI is a solid option if you want to test GPT-4o, Claude 3.5, or Gemini 2.0 from mainland China without paying a cent. The free tier removes all financial risk. But don’t bet your production pipeline on it — 94% uptime and a 56.4/100 composite score tell you this is early-stage software. The limited model selection and lack of pricing roadmap make it a temporary solution, not a long-term platform.
Use it for prototyping, hackathons, or personal projects. For anything customer-facing, wait until they release paid tiers with real SLAs.
FAQ
Q: Does 接口AI require a VPN in China? A: No. It works without VPN from mainland China.
Q: What payment methods are supported? A: Alipay and WeChat Pay. No international credit cards or crypto.
Q: Can I use models outside GPT-4o, Claude 3.5, and Gemini 2.0? A: No. Only those three models are listed in the platform data.
Q: Is there a minimum recharge amount? A: Not specified. The current plan is free with no recharge requirement.
Q: What is the uptime guarantee? A: 94% uptime is reported, with no SLA or refund policy published.
Pricing breakdown
接口AI offers competitive pricing for developers. Here's the breakdown:
| Plan | Price | Quota | Best for |
|---|---|---|---|
| Free | $0/mo | Free trial | 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
3 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 接口AI?
Testing additional relay options