GPT-6 Astra 是 OpenAI 面向高要求端到端任务的旗舰模型。它适用于高级分析、软件工程、深度研究、科学工作和文档创作,在涉及计算机与浏览器使用的长周期智能体任务方面尤为突出。
模态
输入 / 输出价格
每 1M token $10 / $50
上下文窗口
1M
发布时间
2026 年 9 月 4 日
服务商
不同公司托管同一模型。OpenRouter 会根据你选择的路由模式将请求发送至其中一家——Balanced(价格 + 速度)、Nitro(最快)或 Exacto(工具调用准确率最高)。
这是客户实际为该模型支付的平均价格,与各服务商公布的价格并列展示。缓存和折扣意味着实际支付价格往往远低于标价。
性能
吞吐量衡量模型生成文本的速度(每秒 token 数——越高越好)。延迟是完整的往返时间(越低越好)。TTFT 是首 token 时间——即你看到任何内容出现之前需要等待多久(越短越好)。
正常运行时间
正常运行时间是指过去 3 天内至少有一个提供商在响应请求的时间百分比。可用性是指推理成功提供服务的时长百分比。OpenRouter 会持续监控,当某个提供商返回错误时,会自动切换到次优提供商。
基准评测
在标准化评测中的得分。百分比越高越好——排名百分位则显示该模型在 OpenRouter 上所有模型中所处的位置。
应用
向该模型发送最多流量的公开应用。这些应用能很好地反映真实的生产工作负载形态——也暗示了该模型最适合哪些使用场景。
活跃度
该模型随时间变化的 token 用量和请求流量。
探索更多模型
提供商 输入 /M 输出 /M 缓存读取 /M 延迟 吞吐量 正常运行时间 OpenAI Flex $5.00 $25.00 $0.50 2.32s 53 tps 100.00% Azure $10.00 $50.00 $1.00 4.81s 48 tps 99.64% OpenAI $10.00 $50.00 $1.00 3.33s 39 tps 99.96% Azure(美国) $11.00 $55.00 $1.10 4.11s 43 tps 99.86% OpenAI Fast $20.00 $100.00 $2.00 3.35s 54 tps --
吞吐量
54 tok/s
P50,各提供商中最佳
延迟
2.32s
P50,最佳提供商
100.00%
98.85%
过去 3 天的可用性
过去 24 小时的可用性
当上游提供商发生错误时,如果您的请求过滤条件允许,我们可以通过路由到其他健康的提供商来进行恢复。您可以通过 Endpoints API 以编程方式访问各提供商的正常运行时间数据。详细了解我们的负载均衡和自定义选项。
什么是 GPT-6 Astra?
GPT-6 Astra 是 OpenAI 面向高要求端到端任务的旗舰模型。它适用于高级分析、软件工程、深度研究、科学工作和文档创作,在涉及计算机和浏览器使用的长周期智能体任务中尤为突出。
GPT-6 Astra 的价格是多少?
GPT-6 Astra 的定价为输入 token 每百万 $10.00,输出 token 每百万 $50.00;另有独立费率:缓存读取每百万 token $1.00,缓存写入每百万 token $12.50,网络搜索每次调用 $10.00/1K。
GPT-6 Astra 的上下文长度是多少?
GPT-6 Astra 拥有 1,050,000 token 的上下文窗口,最多支持 128,000 个补全 token。
GPT-6 Astra 是否支持工具调用和结构化输出?
支持。GPT-6 Astra 接受 tools 和 tool_choice 参数用于函数调用。它还通过 response_format 中的 JSON schema 支持结构化输出。
tools
tool_choice
response_format
GPT-6 Astra 支持哪些输入和输出?
GPT-6 Astra 接受 PDF、图像和文本等文件作为输入,并返回文本。
哪些提供商提供 GPT-6 Astra 服务?
GPT-6 Astra 在 OpenRouter 上由 2 家提供商提供服务:OpenAI 和 Azure(美国)。请求会被路由到最佳可用提供商,并自动故障转移到其他提供商,你也可以通过提供商路由来固定或排除某些提供商。
GPT-6 Astra 是什么时候发布的?
GPT-6 Astra 于 2026 年 9 月 4 日发布。
GPT-6 Astra is OpenAI's flagship model for demanding end-to-end work. It is suited for advanced analysis, software engineering, deep research, scientific work, and document creation, with particular strengths in long-horizon agentic tasks that involve computer and browser use.
Modalities
In / Out Price
$10 / $50per 1M
Context
1M
Released
Sep 4, 2026
Providers
Different companies host the same model. OpenRouter routes your request to one of them based on the routing mode you pick — Balanced (price + speed), Nitro (fastest), or Exacto (highest tool-calling accuracy).
The average price customers actually pay for this model, next to the prices providers post. Caching and discounts mean the price actually paid is often well below the listed one.
Performance
Throughput is how fast the model writes (tokens per second — higher is better). Latency is total round-trip time (lower is better). TTFT is time-to-first-token — how long before you see anything appear (lower is better).
Uptime
Uptime is the percentage of the past 3 days that at least one provider was responding to requests. Availability is the percentage of time that inference was successfully served. OpenRouter continuously monitors and uses the next-best provider when one returns an error.
Benchmarks
Scores on standardized evaluations. Higher percentages are better — and rank percentile shows where this model lands among all models on OpenRouter.
Apps
Public apps that send the most traffic to this model. Good signal for what real production workloads look like — and a hint at which use cases this model is best suited for.
Activity
Token volume and request traffic to this model over time.
Explore more models
ProviderInput /MOutput /MCache read /MLatencyThroughputUptimeOpenAI Flex$5.00$25.00$0.502.32s53 tps100.00%Azure$10.00$50.00$1.004.81s48 tps99.64%OpenAI$10.00$50.00$1.003.33s39 tps99.96%Azure (US)$11.00$55.00$1.104.11s43 tps99.86%OpenAI Fast$20.00$100.00$2.003.35s54 tps--
Throughput
54tok/s
P50, best across providers
Latency
2.32s
P50, best provider
100.00%
98.85%
Availability over the last 3 days
Availability over the last 24 hours
When an error occurs in an upstream provider, we can recover by routing to another healthy provider, if your request filters allow it. You can access per-provider uptime data programmatically through the Endpoints API. Learn more about our load balancing and customization options.
What is GPT-6 Astra?
GPT-6 Astra is OpenAI's flagship model for demanding end-to-end work. It is suited for advanced analysis, software engineering, deep research, scientific work, and document creation, with particular strengths in long-horizon agentic tasks that involve computer and browser use.
How much does GPT-6 Astra cost?
GPT-6 Astra costs $10.00/M input tokens and $50.00/M output tokens, with separate rates for Cache Read at $1.00/M tokens, Cache Write at $12.50/M tokens and Web Search at $10.00/1K calls.
What is the context length of GPT-6 Astra?
GPT-6 Astra has a 1,050,000 token context window. It supports up to 128,000 completion tokens.
Does GPT-6 Astra support tool calling and structured outputs?
Yes. GPT-6 Astra accepts tools and tool_choice for function calling. It also supports structured outputs via a JSON schema in response_format.
tools
tool_choice
response_format
What inputs and outputs does GPT-6 Astra support?
GPT-6 Astra accepts files such as PDFs, images and text as input and returns text.
Which providers serve GPT-6 Astra?
GPT-6 Astra is served by 2 providers on OpenRouter: OpenAI and Azure (US). Requests are routed to the best available provider, with automatic failover to the others, and you can pin or exclude providers with provider routing.
When was GPT-6 Astra released?
GPT-6 Astra was released on September 4, 2026.