OpenAI发布GPT-6.2 Astra与Sol:面向专业工作的前沿模型,API价格下调约20%OpenAI Launches GPT-6.2 Astra and Sol: Frontier Models for Professional Work at ~20% Lower API Prices


AI free token/AI免费积分

OpenAI发布GPT-6.2 Astra与Sol:面向专业工作的前沿模型,API价格下调约20%

本文来源:https://dev.to/iacompa/openai-launches-gpt-62-astra-frontier-model-for-professional-work-at-20-lower-api-prices-5an6
若有版权问题,请来信告知:13951431913#139.com 或在右方提交评论。

OpenAI 推出了 GPT-6.2 Astra,这是一款面向高强度知识和专业工作的新旗舰模型。该模型在ChatGPT中以及通过API提供,并取代了ChatGPT产品线中的GPT-5.2。

Astra 被定位为高要求工作的默认模型,在 SWE-Bench Pro、GDPval 和多个 GPQA 变体等基准测试中优于 GPT-5.2。与此同时,与 GPT-5.2 相比,它将 API 价格降低了约 20%。

Astra是GPT-5.2的继任者,面向高强度专业工作。根据OpenAI的说法,该模型旨在:

在 STEM 领域更准确,包括编程和物理学

在经济、法律、医疗和金融领域更有帮助

根据内部测试,较少出现幻觉

OpenAI 还报告称,Astra 完成 GDPval 任务的速度比 GPT-5.2 快 1.5 倍。

Astra 被宣传为功能更强大且使用成本更低。对于大规模运行该模型的团队来说,这种组合具有重要意义:基准涵盖专业工作、编程和STEM推理,而价格变动则使相同工作负载的成本降低了约20%。

2026年11月18日 — 本文已更正,以修正GPT-5.2的令牌上下文:应为40万,而非50万。

OpenAI Launches GPT-6.2 Astra and Sol: Frontier Models for Professional Work at ~20% Lower API Prices

本文来源:https://dev.to/iacompa/openai-launches-gpt-62-astra-frontier-model-for-professional-work-at-20-lower-api-prices-5an6
若有版权问题,请来信告知:13951431913#139.com 或在右方提交评论。

OpenAI has launched GPT-6.2 Astra, a new flagship model intended for high-intensity knowledge and professional work. It is available in ChatGPT and through the API, and it replaces GPT-5.2 in the ChatGPT lineup.

Astra is positioned as the default model for demanding work, improving GPT-5.2 on benchmarks such as SWE-Bench Pro, GDPval and multiple GPQA variants. At the same time, it reduces API prices by around 20% versus GPT-5.2.

Astra is the successor to GPT-5.2, aimed at high-intensity professional work. According to OpenAI, the model is designed to be:

More accurate in STEM, including code and physics

More helpful in economics, law, health and finance

Less prone to hallucination, according to internal tests

OpenAI also reports that Astra can solve GDPval tasks 1.5x faster than GPT-5.2.

Astra is presented as both more capable and cheaper to use. That combination is relevant for teams that run the model at scale: the benchmarks cover professional work, coding and STEM reasoning, while the price change lowers the cost of the same workloads by about 20%.

November 18, 2026 — This article was corrected to fix the GPT-5.2 token context: it is 400K, not 500K.

暂无评论,快来发表第一条评论吧!No comments yet. Be the first!

📮 需求咨询📮 Inquiry