This page is a dated editorial pricing snapshot. The rates, trial terms, routing, and model availability below are not independently verified; check N1N’s current terms before purchase.
Pricing Breakdown
N1N does not publish a public per-model rate card in the source record used for this page. The figures below are recorded editorial observations, not a current provider quote. The free tier (free_trial: True) gives you $1 in credit — no credit card required, just registration. That’s enough for roughly 400,000 GPT-4o input tokens or 2 million DeepSeek V3 input tokens. Bottom line: The free trial doesn’t require payment details, so you can test latency and model quality before committing.
| Model | Input (per 1M tokens) | Output (per 1M tokens) |
|---|---|---|
| GPT-4o | $2.50 | $10.00 |
| GPT-4 Turbo | $10.00 | $30.00 |
| Claude 3.5 Sonnet | $3.00 | $15.00 |
| Claude 3 Opus | $15.00 | $75.00 |
| Gemini 2.5 Pro | $1.25 | $5.00 |
| DeepSeek V3 | $0.27 | $1.10 |
| These rates reflect 2026 pricing and change. N1N charges in USD, which means Chinese users pay a currency conversion overhead of roughly 1-2% depending on your bank. |
Hidden Fees and Batch Discounts
N1N does not charge per-request overhead fees. Each API call bills exactly for the tokens consumed. No minimum top-up amount is specified (Min recharge: Not specified), so you can recharge any amount via Alipay, WeChat Pay, or USDT. Batch discounts are absent from the current pricing model. Unlike some relay stations that offer 5-10% off for prepaid balances above ¥500, N1N charges flat per-token rates regardless of volume. Key takeaways:
- No per-request overhead or hidden fees
- No batch or volume discounts available
- USD pricing adds 1-2% currency conversion cost for Chinese users
Cost Comparison vs Direct API
For users inside China, direct API access to OpenAI or Anthropic requires a VPN. That adds $5-15/month for a reliable VPN service plus latency overhead. N1N’s 162ms average latency (Latency: 162ms) is faster than most VPN-routed connections, which typically sit at 300-500ms.
| Cost Factor | Direct API + VPN | N1N |
|---|---|---|
| Monthly VPN | $5-15 | $0 (included) |
| GPT-4o per 1M tokens | $2.50 + VPN | $2.50 |
| Setup friction | High (card, VPN config) | Low (register, pay) |
| Uptime guarantee | None | 99.3% |
| The breakeven point is roughly 500,000 tokens per month. Below that, direct API with a cheap VPN is cheaper. Above that, N1N wins on convenience and reliability. |
Pros & Cons
Pros
- 24 CN2 global proxy nodes — fastest overseas routing for Chinese users
- 99.3% uptime (Uptime: 99.3%) — production-ready
- Free trial with $1 credit, no credit card required
- Supports Alipay and WeChat Pay for Chinese users
- 28+ model variants available (Models: GPT-4o, GPT-4 Turbo, Claude 3.5 Sonnet, Claude 3 Opus, Gemini 2.5 Pro, DeepSeek V3) Cons
- USD pricing more expensive for Chinese users due to currency conversion
- Registration required — no anonymous access
- Fewer Chinese-focused features compared to domestic relay stations
- No batch discounts or volume pricing
Verdict
N1N is the best option for Chinese developers who need production-grade API access to Western LLMs without VPN hassle. The 162ms latency and 99.3% uptime beat any VPN-based setup. The free trial lets you verify performance before spending. If you process over 500,000 tokens monthly, the cost is competitive with direct API + VPN. For smaller volumes, the convenience premium is worth it.
FAQ
Does the N1N free trial require a credit card?
No. Registration is free and you get $1 in credit without entering payment details.
What’s the minimum recharge amount in CNY?
Not specified. You can recharge any amount via Alipay, WeChat Pay, or USDT.
Does N1N offer batch discounts for high-volume users?
No. N1N charges flat per-token rates with no volume or batch discounts currently available.
Is N1N accessible from China without a VPN?
Yes. N1N has 24 CN2 global proxy nodes optimized for Chinese users, eliminating the need for a separate VPN.
What models does N1N support?
N1N supports GPT-4o, GPT-4 Turbo, Claude 3.5 Sonnet, Claude 3 Opus, Gemini 2.5 Pro, and DeepSeek V3 — 28+ model variants total.
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.
