Publish time
2026/8/27Model Series
千问Input type
Output type
Input Price
¥1 / 1M tokensOutput Price
¥3 / 1M tokensCache Write Price
¥1.25 / 1M tokensCache Read Price
¥0.1 / 1M tokensContext Window
1,000,000Max Output Length
128,000Qwen3.8-Flash 是千问最新推出的多模态大模型,兼具强大的理解与生成能力和出色的响应速度。模型原生支持百万级上下文窗口,能够一次性处理超长文档、代码仓库和复杂对话。在编程辅助、智能体协作、图文理解等场景中表现尤为出色——无论是自动修复代码、操作桌面应用,还是分析图表与长视频,都能给出准确、高质量的结果。同时兼容 OpenAI 与 Anthropic 主流接口协议,可无缝接入 Claude Code、Codex 等开发者工具,轻松构建高并发应用与智能工作流。凭借优异的性能与极具竞争力的推理成本,Qwen3.8-Flash 是开发者和企业在 AI 应用中兼顾效果与效率的理想选择。
Zhinao API routes requests to the best-fit provider and automatically fails over to the one with highest availability.
TTFT
1.60s
Throughput
51.07tps
Uptime
100.00%
Provider Model
alibaba/qwen/qwen3.8-flash
Supported Parameters
Recent Uptime
Reasoning
Toggleable
Supported Response Formats
Total Context
1,000,000
Max Output
128,000
Input Price
¥1 / 1M tokens
Output Price
¥3 / 1M tokens
Cache Write
¥1.25 / 1M tokens
Cache Read
¥0.1 / 1M tokens
Compare different providers across Zhinao API
61.41 tok/s
1.70 s
Uptime for qwen/qwen3.8-flash across all providers
Zhinao API normalizes requests and responses across providers for you
import OpenAI from "openai";
const client = new OpenAI({
baseURL: "https://api.360.cn/v1",
apiKey: process.env.ZHINAO_API_KEY,
});
const response = await client.chat.completions.create({
model: "qwen/qwen3.8-flash",
messages: [
{ role: "user", content: "Hello, how are you?" }
],
temperature: 0.7,
max_tokens: 1000,
});
console.log(response.choices[0].message.content);