OpenAI has the #2 model in the Intelligence Index with the release of GPT-6.2 Astra and GPT-6.2 Sol November 17, 2026
OpenAI发布了GPT-6.2 Astra和GPT-6.2 Sol。
6 reports9 sources
Reports 6
OpenAI rolls out GPT-6.2 Astra and Sol as Microsoft and Google launch new AI capabilities
OpenAI is rolling out GPT-6.2 Astra and Sol to ChatGPT Work and Codex users on Plus, Pro, Enterprise and Edu plans, and to developers through its API; both models support a 1 million token context. Microsoft has made two coding models available in GitHub Copilot, Copilot Fusion and Foundry, while Google has launched free Hydrogen-powered Gemini Tasks in six countries, including tasks requiring a phone call or text.
OpenAI launches GPT-6.2 Astra and Sol for ChatGPT Work, Codex users and developers
GPT-6.2 comes in Astra, a flagship model, and Sol, an everyday model. The report says professional work and coding capabilities have improved while token costs have fallen substantially. Both are now available to ChatGPT Work and Codex users and developers.
2 more sources reported this
- IT之家 科技新闻
- Tibo (@thsottiaux)
OpenAI releases GPT-6.2 Astra and Sol for professional projects and everyday work
OpenAI has released GPT-6.2 Astra for complex professional projects and GPT-6.2 Sol for everyday work. OpenAI says Astra scored 5.2 percentage points higher than GPT-6 Astra on an agent benchmark at less than one-third the per-task cost. Both models are available to eligible ChatGPT Work and Codex users and through the API, but have not been added to ChatGPT’s regular chat mode.
2 more sources reported this
- IT之家 科技新闻
- Tibo (@thsottiaux)
OpenAI releases GPT-6.2 Astra and Sol; Astra ranks second in evaluation, while coding agent results lag
Artificial Analysis ranks GPT-6.2 Astra second in its Intelligence Index, up 5 points from GPT-6 Astra. Astra costs $8 per million input tokens and $40 per million output tokens, 20% less than its predecessor. In the Coding Agent Index, Astra is flat with its predecessor, while Sol scores 3 points below GPT-6.1 Sol.
Artificial Analysis updates Endpoint Accuracy Index; 24 of 103 endpoints score below self-hosted references
Artificial Analysis launched Endpoint Accuracy Index v1.1 with coverage for GLM-5.3 Flash and Muse Spark Open. Of 103 serverless endpoints tested, 24 scored statistically below self-hosted references. More repeats narrowed the average confidence interval for its long-context test; raw scores from that test are not directly comparable across versions.
Artificial Analysis updates its model capability index with two new benchmarks
Artificial Analysis released Intelligence Index v5 with new benchmarks for real-world coding tasks and scientific research workflows in a terminal. Private test sets or held-out answers now underpin 60% of the index. In its results, Claude Opus 5.5 (max) leads with 58 points, followed by GPT-6 Astra (max) with 57. The organization also launched Progress Index v1 to compare model capability over time.
1 more source reported this
- IT之家 科技新闻
SourceArtificial Analysis 模型评测 ↗MarkTechPost AI 研究报道 ↗Simon Willison AI 实践 ↗IT之家 科技新闻 ↗IT之家 科技新闻 ↗Artificial Analysis 模型评测 ↗Artificial Analysis 模型评测 ↗Latent Space AI 报道 ↗ginobefun (@hongming731) ↗