SyncAI.news, a Varaisys broadcasting
OpenAI’s GPT-6.2 Astra and Sol arrive in GitHub Copilot with smarter coding and better code reviews
MA

Mint: AI

· 1 min read

IndiaMint: AI

OpenAI’s GPT-6.2 Astra and Sol arrive in GitHub Copilot with smarter coding and better code reviews

OpenAI has launched GPT-6.2 Astra and GPT-6.2 Sol in GitHub Copilot, bringing improved coding, reasoning and code review capabilities with better token efficiency.

OpenAI’s GPT-6.2 Astra and GPT-6.2 Sol are now available in GitHub Copilot, bringing improvements in agentic coding, code reviews and complex software development tasks. The models were announced today, 18 November, to enable developers to pinpoint defects in their code more precisely, solve multi-step tasks and save tokens, according to OpenAI's official announcement.

It is being rolled out to GitHub Copilot Pro, Max, Business and Enterprise subscribers. Both models achieved improved performance in initial code review tests when compared with their predecessors, GitHub says. GPT-6.2 Astra is optimized for advanced coding and terminal-based tasks, whereas GPT-6.2 Sol is geared towards intricate agentic workflows and comprehensive code reviews.

GPT-6.2 Astra focuses on advanced coding and professional tasks

GPT-6.2 Astra is built for certain professional tasks, such as software development, scientific research or computer-based tasks, according to OpenAI's announcement on 17 November. The model can create more well-designed presentations, spreadsheets and documents and adhere to previously established models, the company said.

OpenAI also claimed to have enhanced coding capabilities, including website development, gaming and 3D design. On the Agents' Last Exam benchmark, which evaluates multi-step professional tasks, GPT-6.2 Astra scored 5.2 percentage points higher than GPT-6 Astra. It also set a new record on Terminal-Bench-Science 0.1, improving by 9 points over its predecessor.

GPT-6.2 Sol improves everyday coding and agentic workflows

GPT-6.2 Sol aims to be a more efficient model for general professional and technical tasks. It enhances reasoning, coding and multi-step workflows without increasing operational costs.

Pragya Singha Roy

Original source

This story was published by Mint: AI. SyncAI.news shows a preview; the complete article is on the publisher's site.

Read the full story on livemint.com

Similar News