Unverified50% confidenceBenchmarkExact time
hyperframes-registry (heygen-com/hyperframes) improved from 0.43 to 0.98 on Sonnet and from 0.62 to 0.97 on Opus.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/19/2026
First Seen
Valid until: 1/17/2027
Sources
Related Entities
Related Claims
UnverifiedSonnet 5.5's reasoning tiers are divided into low/medium/high/x-high/Max, with token usage increasing only 5% to 8% from low to x-high, but jumping approximately 1500% (15×) from x-high to Max.63% similarUnverifiedIn test one (landing page design), the creator awarded the win to Sonnet 5.5, on the grounds that Sonnet's above-the-fold looked better and replacing its fake jar with real assets would cost very little, while Opus ran for 52 minutes versus Sonnet's 28 minutes, with Opus costing nearly twice as much as Sonnet.62% similarVerifiedClaude Sonnet 5.5 achieved 70.6% on the Terminal benchmark, compared to only 10.3% for the previous generation.61% similarUnverifiedThe scanner's live AI (using Claude Sonnet 4.5 by default with GPT-4o mini as backup) had a rank correlation of +0.55 between its lift scores and past stock price gains, a correlation of only -0.06 with 5-day forward returns, and the correlation between past gains and future performance was approximately 0.00.61% similarUnverifiedThe two-stage classifier runs on top of Claude Sonnet 5.5, Opus 5 and 5.5, and Fable 5 and 5.1, and Anthropic says its false positive rate is below 0.1%.60% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1065268API
curl https://kongchang.com/api/v1/knowledge/claims/1065268MCP
get_claim(id=1065268)