Unverified50% confidenceBenchmarkExact time
OpenAI 6.2 Astra scores 9 percentage points higher than the previous generation 6 Astra on Terminal Bench Science.
1
Sources
50%
Confidence
Short-term (~14 days)
Relevance
11/18/2026
First Seen
Valid until: 12/2/2026
Sources
OpenAI 6.2 Astra Hands-On Review Gone Wrong: Impressive Benchmarks, But Developers Say It's Unusable
youtubeTheo - t3․gg11/18/2026
Related Entities
Related Claims
UnverifiedIn Terminal Bench Science at maximum reasoning intensity, Astra ranked first with 68.1%, and it is officially recommended for the most difficult scientific research tasks.76% similarUnverifiedOn the Bench 2.0 benchmark, an internal model scored 0.875 while Astra scored 0.594; the internal model's lowest reasoning tier scored 0.758, already surpassing Astra's result when running at full effort.70% similarUnverifiedOpus 5.5 is not first across the board: on Automation Bench it scored 40, slightly below Astra's 41.4, and on Terminal Bench Science it trails GPT-6 Astra.70% similarUnverifiedOpenAI 6.2 Astra scored 78.3 on the Agent's Last Exam benchmark, 5.5 percentage points higher than Claude Fable 5.2.69% similarVerifiedOn Terminal Bench 4.0, Opus 5.5 scored 66.4%, Astra scored 57.9%, Fable 5.1 scored 55.8%, and the previous generation Opus 5 scored 52.3%.66% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1185491API
curl https://kongchang.com/api/v1/knowledge/claims/1185491MCP
get_claim(id=1185491)