GPT-6.2 Astra tops browser use at 93.0 but drops to 70.8 on agentic tasks, review finds
bindureddy · x · 2026-11-18
Bindu Reddy's review of OpenAI's GPT-6.2 Astra argues the company is sacrificing performance for cost:
- Agentic tasks nerfed: score dropped from GPT-6 Astra's 75.6 to 70.8
- Excellent at browser use: 93.0, tied for the top score
- Very token-efficient: $1.941 per successful task, far below Claude Fable 5.2's $9.850, though above GPT-6.1 Sol's $0.823
- Overall 84.9, tying Claude Fable 5.2 and beating Gemini 4 Carbon (83.5)
- GPT-6.2 Sol is even more disappointing: 80.3 overall, below Sol 6.1's 81.4, with agentic at just 63.6
The benchmark table suggests Astra is positioned as a cheap, browser-strong model at the expense of agentic capability.
More from Models
- GPT-6.2 Astra and Sol roll out as testers push back on the hate wave — jeffintime · 2026-11-18
- GPT 6.2 Astra and Sol Take 4th and 5th on sudo L7 Coding Benchmark, Still Behind Latest Claude — echen · 2026-11-18
- Why did OpenAI ship a dud? Ex-engineer suspects cost-cut bet on RL backfired — doodlestein · 2026-11-18
- No single model matches Claude, says developer as Anthropic 'hits escape velocity' — _arohan_ · 2026-11-18
- Hands-on with GPT-6.2: Ignore the hate, it's a cheap computer-use workhorse — nickbaumann_ · 2026-11-18
- Claude leads Arena agent leaderboard with nearly 2x praise-to-complaint ratio — thursdai_pod · 2026-11-18