金石APIAuthenticity, speed and price analysis
About the site
金石 API 是面向开发者与 AI 工具用户的多模型 API 聚合平台,尽力提供低价而稳定高效的API服务。通过统一入口接入多种 AI 模型,自由选择模型和渠道分组,并在控制台查看调用记录、额度消耗与模型响应信息。从日常对话、辅助编程到应用开发,让模型接入与调用管理更方便、费用更清楚。
Service details
- Minimum top-up¥1
- Sign-up bonusNot provided
- Support responseNot provided
- Payment methodsAlipay, WeChat Pay
- China invoiceConditional
- International invoiceNot verified
- RefundsNot verified
- PromotionsNot provided
(seconds)
coding session
availability
Comparison summary
GPT-4o Mini on 金石API: a coding session costs ¥0.65, ranking 1 of 1, the cheapest relay (median ¥0.65).
Authenticity
Model consistency
Output distribution for the same requests on this relay and on the official API · higher is better
Speed
Measured by time to first token (seconds) and output speed (tokens / second)
Time to first token
Seconds until the first token · lower is better
Price
Groups and models
| Group | Model | Availability | Input price | Output price | Pricing unit |
|---|---|---|---|---|---|
| GPT-Azure | GPT-4o Mini | Not probed yet | ¥0.30 | ¥1.20 | Per million tokens |
Cost per coding session
GPT-4o Mini · 2M input and 40k output tokens (¥) · split by token type · lower is better
Prices: cache hit, input and output
¥ / 1M tokens
Cache
Coding requests repeat much of their context, so the cache hit rate largely decides what you actually pay. Drag the hit rate and the costs and ranking below are reweighted with cache prices.
Cache-weighted session cost
GPT-4o Mini · of 2M input tokens, the hit share is billed at the cache price and the rest at the input price, plus 40k output (¥) · lower is better
Stability
Availability
Request success rate, last 24 hours · higher is better
Status of each model, last 24 hours
One cell per hour · green ok, yellow delayed, red failed
| Model | Group | Availability | Last 24 hours | First token | Samples |
|---|---|---|---|---|---|
| GPT 4.1 | GPT-Azure | — | — | — | |
| GPT 4.1 Mini | GPT-Azure | — | — | — | |
| GPT 5 Nano | GPT-Azure | — | — | — | |
| GPT 5 | GPT-Azure | — | — | — | |
| GPT-4o | GPT-Azure | — | — | — | |
| Claude Haiku 4.5 | Claude-Max-API蒸馏 | — | — | — | |
| GPT-4o Mini | GPT-Azure | — | — | — | |
| Gemini 4 Argon | Gemini | — | — | — | |
| DeepSeek V4 Pro | DeepSeek-华为 | — | — | — | |
| Claude Sonnet 5 | Claude-CC4 | — | — | — | |
| Kimi K3 | Kimi-Cyber | — | — | — | |
| GPT 5.6 Terra | gpt-enterprise | — | — | — | |
| Claude Fable 5 | Claude-Max-API蒸馏 | — | — | — | |
| GPT 5.6 Sol | GPT-不降智2 | — | — | — | |
| Claude Opus 5 | Claude-CC4 | — | — | — | |
| Claude Opus 4.7 | Claude-CC4 | — | — | — | |
| Claude Opus 4.8 | Claude-CC4 | — | — | — | |
| Claude Opus 4.6 | Claude-CC4 | — | — | — | |
| Claude Sonnet 4.6 | Claude-CC4 | — | — | — | |
| Gemini 3.8 Flash | Gemini | — | — | — | |
| GLM 5.3 | Cyber-Pro | — | — | — | |
| Claude Sonnet 4.5 | Claude-Max | — | — | — | |
| Claude Opus 4.5 | Claude-Max | — | — | — | |
| Gemini 2.5 Flash | Gemini | — | — | — |
FAQ
How much does GPT-4o Mini cost on 金石API?
Input ¥0.3 and output ¥1.2 per 1M tokens; at 2M input and 40k output tokens, a coding session costs about ¥0.65, an API rate of 0.30.
How does the cache hit rate affect price?
金石API has not published a cache price for this route, so cache hits are billed at the input price.Measured hit rates will replace this assumption once available.
Which models does 金石API support?
Folkbench currently lists 24: Claude Sonnet 5, Claude Fable 5, GPT 5.6 Sol, GPT 5.6 Terra, Claude Opus 5, Kimi K3, DeepSeek V4 Pro, Claude Opus 4.8, Claude Opus 4.7, Claude Opus 4.6, Claude Sonnet 4.6, Gemini 3.8 Flash, GLM 5.3, Claude Haiku 4.5, Gemini 4 Argon, GPT-4o, GPT-4o Mini, Claude Sonnet 4.5, Claude Opus 4.5, Gemini 2.5 Flash, GPT 4.1, GPT 4.1 Mini, GPT 5 Nano, GPT 5.
What does the data security level mean?
Folkbench rates relays from 0 to 4 based on public policies and verifiable evidence; 金石API is currently level 0. The check covers leaks in responses and error messages only, and does not prove data retention or training policies.
How often is this data updated?
Availability and time to first token use a 24-hour window, and this page syncs live data every 15 minutes.