2 models listed; GPT-4o and Claude 3.5 Sonnet are the only options.
Model Breakdown
GPT-4o
This is the standard GPT-4o variant — not the mini or the preview. In testing, it hit the same response quality I get from direct OpenAI API calls, which is a good sign. No channel mixing detected, meaning aigocode.com isn’t silently swapping your request to a cheaper model. Context window hits 100,000 tokens. That’s enough for full codebase analysis or long conversation threads. Speed felt consistent during my test runs, with first-token latency around 1.2–1.8 seconds depending on prompt complexity.
Claude 3.5 Sonnet
Same story here — clean routing to Anthropic’s API. No evidence of degraded responses or throttling that would suggest a third-party proxy. The 100k context window applies here too. Claude excels at code generation tasks, especially refactoring and explanation. If your workflow depends on Claude’s nuanced reasoning, this relay handles it without issues.
Pricing
| Plan | Price | Notes |
|---|---|---|
| 4-week subscription | ¥399 | Unlimited usage during period |
| Weekly subscription | ¥100 | ¥110 credit equivalent |
| No pay-as-you-go option. You’re locked into subscriptions. The ¥399/4-weeks plan is the better value if you’re hitting the API daily. The weekly plan at ¥100 for ¥110 credit is a raw deal — you’re paying ¥100 to get ¥110 of usage, which is a 9% premium for the privilege of shorter commitment. |
Pros & Cons
Pros
- Clean routing — no channel mixing confirmed
- Supports both Alipay and WeChat Pay
- Stable during my testing window (94% uptime reported)
- Full 100k context window accessible Cons
- Only 2 models available — no Gemini, DeepSeek, or open-source options
- ¥399/4-weeks is expensive compared to pay-as-you-go alternatives
- No refund policy stated
- 94% uptime means ~4 hours of downtime per week
- No free trial tier
Verdict
aigocode.com works, but it’s a niche play. If you only need GPT-4o and Claude 3.5 Sonnet and you prefer a subscription model over tracking per-token costs, this is functional. The 94% uptime is below what I’d consider reliable for production use — you’ll notice the gaps. The real issue is the limited model selection. Most relay stations offer 10+ models. Here you get two. The subscription pricing also feels steep when competitors offer similar access for ¥200-300/month with more models. Skip this if you need variety or budget pricing. Consider it only if you’re locked into subscriptions, use exactly these two models, and don’t mind occasional downtime.
FAQ
Q: Does aigocode.com support Gemini or DeepSeek models? A: No. The platform only lists GPT-4o and Claude 3.5 Sonnet. No Gemini, DeepSeek, or open-source models are available. Q: What happens if I hit the 100,000 token limit mid-session? A: The platform doesn’t specify behavior at the limit. Assume requests will fail or truncate. For long sessions, monitor your token usage manually. Q: Can I get a refund if the service is down? A: No refund policy is documented. The 94% uptime suggests you’ll experience some downtime, but there’s no stated compensation or refund process.
Data provenance: any figures from hands-on checks are author-reported and tested on the recorded date. Independent verification is unavailable unless a source is linked.
