Chat, reasoning, images, long context — per-call pricing, clearly labeled. Every credit you spend, accounted for.
Why HeySmart
Billing
Every model's billing mode is labeled in the price list — you know the cost before sending a request.
Fits: daily chat, light integrations, budget-sensitive workloads
Fits: long documents, long agent runs, bulk production calls
Top-up rate: 1 CNY = 10 credits. The more you top up, the more bonus credits — 50 credits free on signup, no card required
Models
The model name is the call name — just pass it as the model parameter. No per-vendor accounts, no juggling multiple keys.
Plans
Pick a plan that matches your pace; upgrade or downgrade anytime, and top up credits whenever you run short.
Get started
Sign up in 30 seconds and get 50 free credits instantly — no credit card required.
One click in the console; one key for every model capability you need.
OpenAI SDK compatible — change two lines of config, zero changes to your code.
FAQ
Still have questions? Check the answers below or reach out to our support team.
HeySmart is a unified AI API gateway. With a single key you can call major mainstream LLMs through the OpenAI SDK — we handle routing, rate limiting, billing and reliability for you, so there's no need to integrate with each vendor or maintain multiple SDKs and bills.
Yes. Sign-up grants 50 free credits with no payment method required — enough to try chat, reasoning, image generation and every other capability hands-on. Top up only when you run out.
Two modes: per-call models charge a fixed number of credits per request (e.g. 1 credit/call, visible in the console price list before you send), regardless of tokens or characters; pay-as-you-go models are settled precisely on actual input/output tokens. No hidden surcharges, ever.
1 CNY = 10 credits. Top-up tiers include bonus credits: top up ¥10 and get 115 credits (100 + 15 bonus), ¥50 gets 565, ¥100 gets 1130 — the more you top up, the more bonus you get.
Model families include mainstream LLMs such as Tongyi Qianwen (Qwen), Zhipu (GLM), Kimi and DeepSeek, covering chat, long-context and more, with new models onboarded continuously. See the console price list for the full catalog and live prices.
No. Billing happens only when the upstream returns valid content. Network timeouts, upstream errors and empty responses are not charged, and any pre-deducted credits are automatically refunded to your account.
Top-up credits and bonus credits never expire. Periodic credits included in a plan reset each subscription cycle and do not accumulate across cycles.
After signing up, go to Console → API Keys and generate one with a single click. You can create multiple keys for per-project management and isolation.
Credits are non-refundable once delivered; we recommend topping up based on actual usage. Top-up credits never expire and can be used for API calls or plan subscriptions at any time.
Yes. Enterprise customers can contact support for custom routing, private deployment and dedicated SLAs; high-volume customers can request bulk discounts.
No card required. 50 credits is enough to try chat, reasoning and image generation hands-on.