---
format: "aidr-story-markdown/v1"
id: "7f85bfe669e0c883a518e0087119e95ea25398355fca6c2d5c24e141a2bd066b"
canonical_url: "https://aidr.today/7f85bfe6?lang=en"
title: "GPT-6.2 Astra Ties Grok 4.7 and Trails Claude in Independent Document Tests"
lang: "en"
requested_lang: "en"
available_langs: ["en","vi"]
translation_fallback: null
fallback_fields: []
published_at: "2026-11-21T20:21:29.000Z"
category: "Models"
topics: ["openai","llm"]
source_urls: ["https://marketbrief.now/ai/gpt-62-astra-ties-grok-47-and-trails-claude-in-independent-document-test-44f7657c","https://huggingnews.com/ai/gpt-62-astra-ties-grok-47-and-trails-claude-in-independent-document-test-44f7657c","https://the-decoder.com/gpt-6-2-astra-and-sol-get-to-the-point-faster-but-score-lower-on-transparency-than-claude/"]
summary: "OpenAI's GPT-6.2 Astra model performs roughly equal to Grok 4.7 and falls significantly behind Claude in independent document, spreadsheet, and slide benchmarks. The GPT-6.2 series also includes the Sol model for general work and coding, which is available on NEAR AI Cloud."
---

# GPT\-6\.2 Astra Ties Grok 4\.7 and Trails Claude in Independent Document Tests

> [Open the canonical story](<https://aidr.today/7f85bfe6?lang=en>)

**Published:** 2026-11-21T20:21:29.000Z
**Category:** Models
**Topics:** openai, llm

## Summary

OpenAI's GPT\-6\.2 Astra model performs roughly equal to Grok 4\.7 and falls significantly behind Claude in independent document, spreadsheet, and slide benchmarks\. The GPT\-6\.2 series also includes the Sol model for general work and coding, which is available on NEAR AI Cloud\.

## Sources

- [Story source](<https://marketbrief.now/ai/gpt-62-astra-ties-grok-47-and-trails-claude-in-independent-document-test-44f7657c>)
- [Story source](<https://huggingnews.com/ai/gpt-62-astra-ties-grok-47-and-trails-claude-in-independent-document-test-44f7657c>)
- [Story source](<https://the-decoder.com/gpt-6-2-astra-and-sol-get-to-the-point-faster-but-score-lower-on-transparency-than-claude/>)

