ainewsblitz.com

Breaking

xAI Grok 4.6 Extra High leads CursorBench 3.2 at 70.8% for $2.81 per task

  • Foundation Models
  • Software Dev & Coding
  • AI Agents

xAI’s Grok 4.6 Extra High now sits first on Cursor’s CursorBench 3.2 leaderboard at 70.8%, at an average $2.81 per task—about six times cheaper than Anthropic’s Fable 5 Max at a nearly identical score. The result, published on Cursor’s public evals table, puts the extra-high setting a fraction of a point ahead of Fable 5 Max (70.5%, $17.32) and Claude Opus 5 Max (70.0%, $8.23), and further ahead of OpenAI’s GPT-5.6 Sol Max (67.2%, $5.69).

Continue reading

The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.

$20
Read this article
$29/month
Unlimited — all 27,502 articles, the full archive, and comprehension quizzes
Save 72%
$98/year
≈ $8.17/month
Unlimited, billed once a year