GPT-5.6 LunaAPI
GPT-5.6 fast tier — high-volume, low-latency execution at mini-class cost.
輸入價格
$0.07
per 1M tokens
輸出價格
$0.42
per 1M tokens
Prompt 快取
給穩定 prefix 加 cache_control，命中按 input 價的一小部分計費。cache_write 按 input 原價計費。
按 input × 0.2 自動派生
引數
- 模型 ID
gpt-5-6-luna- 端點
POST /v1/chat/completions- 類別
- Text
- 供應商
- OpenAI
- 能力
- functionstreaminglong-context
無縫相容 OpenAI SDK
把 OpenAI SDK 的 base_url 指向 api.kunavo.com/v1，換 model id 即可。流式、工具呼叫、SDK 行為完全一致。
檢視文件GPT API pricing guide立即試用
在 /app/keys 建立 API key，複製為 KUNAVO_API_KEY，然後任選下面一段程式碼執行。
curl https://api.kunavo.com/v1/chat/completions \
-H "Authorization: Bearer $KUNAVO_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "gpt-5-6-luna",
"messages": [
{"role": "user", "content": "Hello, GPT-5.6 Luna"}
],
"stream": false
}'Kunavo 與直連 OpenAI 的差異
同一個 GPT-5.6 Luna，同樣的輸出。閘道改變的是帳號、SDK 與帳單——以下是如實對照。
| Kunavo | 直連 OpenAI | |
|---|---|---|
| 價格（per 1M tokens） | $0.07 / $0.42 (−65%) | $0.20 / $1.20 |
| 帳號 | 一個 Kunavo 帳號，2 分鐘取得 Key，最低儲值 $10 | 需要 OpenAI 帳號，並另外設定其計費 |
| SDK | 繼續用 OpenAI SDK——改 base_url，model 填 "gpt-5-6-luna" | OpenAI 官方 SDK，或其提供的 OpenAI 相容層 |
| 同一把 Key 還能呼叫 | Claude、Gemini、GPT 以及影像、影片、音訊模型 | 僅 OpenAI 自家模型 |
| 計費 | 一個 Stripe 餘額，用多少付多少，餘額不過期；失敗請求不計費 | 每家廠商一張獨立帳單 |
| 速率限制 | 共享閘道容量，無帳號層級的合約限額 | OpenAI 自家級距限額，可依用量或合約調升 |
若你需要合約級速率限制、企業 SLA，或首日就要用到廠商獨有功能，直連 OpenAI 較合適。若你想要一把 Key、一張帳單，並在所有廠商上取得更低單價，就用 Kunavo。
FAQ
How much does GPT-5.6 Luna cost?
On Kunavo, GPT-5.6 Luna is $0.07 per 1M input tokens and $0.42 per 1M output tokens — about 65% under OpenAI's official price. Failed requests are never billed, and it's pay-as-you-go — you only pay for successful calls.
Can I call GPT-5.6 Luna with the OpenAI SDK?
Yes. Set base_url to https://api.kunavo.com/v1, pass your Kunavo key, and set model to "gpt-5-6-luna". Requests and responses are OpenAI-compatible.
What endpoint does GPT-5.6 Luna use?
GPT-5.6 Luna is called via POST /v1/chat/completions on Kunavo's OpenAI-compatible API.
What is GPT-5.6 Luna good for?
GPT-5.6 Luna is a text model from OpenAI, with support for function, streaming, long-context.
What is different from calling OpenAI directly?
The model is the same, at about 65% under OpenAI's list price. What changes is around it: one key and one Stripe balance instead of a OpenAI account with its own billing, the OpenAI SDK with base_url swapped instead of a provider-specific SDK, and the same key also reaching Claude, Gemini, GPT, image, video and audio models. The tradeoff is that gateway capacity is shared — if you need a contractual rate limit or an enterprise SLA on GPT-5.6 Luna specifically, go direct to OpenAI.
Is GPT-5.6 Luna cheaper on Kunavo?
Yes — GPT-5.6 Luna is about 65% under OpenAI's official list price on Kunavo, with no subscription and no minimum monthly spend. You pay per token from a $10 top-up, failed requests are never billed, and the balance never expires.