OpenAI: GPT-6 Astra Pro (batch)

Text input Image input File input Text output
Author's Description

GPT-6 Astra Pro is the same underlying model as [GPT-6 Astra](https://openrouter.ai/openai/gpt-6-astra), served with `reasoning.mode` set to `pro` for higher-quality responses on complex tasks. Learn more in OpenAI's docs: https://developers.openai.com/api/docs/guides/reasoning#reasoning-mode

Key Specifications
Cost
+$$$$$
Context
1M
Released
Sep 04, 2026
Speed
Ability
Reliability
Supported Parameters

This model supports the following parameters:

Tool Choice Response Format Include Reasoning Reasoning Tools Max Tokens Structured Outputs Seed
Features

This model supports the following features:

Reasoning Tools Response Format Structured Outputs
Performance Summary

GPT-6 Astra Pro, released on September 4, 2026, is a specialized version of GPT-6 Astra optimized for complex tasks through its `reasoning.mode` set to `pro`. The model demonstrates competitive response times, ranking in the 40th percentile across benchmarks. However, it is positioned at a premium price point, ranking in the 2nd percentile for cost. A significant strength is its exceptional reliability, achieving a 100% success rate across all benchmarks, indicating consistent and usable responses. In terms of performance, GPT-6 Astra Pro excels in tasks requiring precision and logical deduction. It achieved perfect accuracy in both Hallucinations (Baseline) and Reasoning (Baseline), notably being the most accurate and fastest at its price point in these categories. Its Instruction Following capabilities are strong at 92.0% accuracy, placing it in the 96th percentile. While its General Knowledge (96.5%) and Ethics (99.0%) scores are high, their percentile rankings (38th and 42nd respectively) suggest a competitive but not leading position in these areas relative to other models. Coding and Mathematics performance are solid, with accuracies of 93.0% in both. The model's primary weakness lies in its premium pricing across all categories, which may limit its cost-effectiveness for certain applications despite its high performance.

Model Pricing

Current Pricing

Feature Price (per 1M tokens)
Prompt $5
Completion $25
Input Cache Read $0.5
Input Cache Write $6.25
Web Search $10000

Price History

Available Endpoints
Provider Endpoint Name Context Length Pricing (Input) Pricing (Output)
OpenAI
OpenAI | openai/gpt-6-astra-pro-20260903 1M $5 / 1M tokens $25 / 1M tokens
Azure
Azure | openai/gpt-6-astra-pro-20260903 1M $10 / 1M tokens $50 / 1M tokens
OpenAI
OpenAI | openai/gpt-6-astra-pro-20260903 1M $10 / 1M tokens $50 / 1M tokens
Azure
Azure | openai/gpt-6-astra-pro-20260903 1M $11 / 1M tokens $55 / 1M tokens
OpenAI
OpenAI | openai/gpt-6-astra-pro-20260903 1M $20 / 1M tokens $100 / 1M tokens
Benchmark Results
Benchmark Category Reasoning Strategy Free Executions Accuracy Cost Duration
Other Models by openai