Unverified50% confidenceBenchmarkExact time
Claude Sonnet 5.5 still leads on all three benchmarks Terminal Bench 4.1, OS World and Terminal Bench Science 0.1, and Terra also surpassed Hydrogen on one of them.
1
Sources
50%
Confidence
Short-term (~14 days)
Relevance
11/15/2026
First Seen
Valid until: 11/29/2026
Sources
Gemini 4 Hydrogen Released: A Lightweight AI Option Focused on Speed and Cost-Effectiveness
youtubeaverage:dev11/13/2026
Related Entities
Related Claims
UnverifiedUnder the highest difficulty settings of Terminal Bench 4.0, Sonnet 5.5 outperforms Opus 5.5.73% similarUnverifiedOn multiple benchmarks (HLE Diamond, AA Omniscience, Terminal Bench, Frontier Code, etc.) Fable 5.2 leads almost across the board, but on AA Briefcase Sonnet 5.5 ranks first, on Automation Bench Carbon ranks first, on GDP.pdf Muse Wonder leads, and on Terminal Bench 4.0 Kimi K 3.1 Pro ranks first.67% similarUnverifiedOn the same Terminal Bench, Claude Sonnet 5.5 scores 63.9, ahead of Hydrogen's 57%; on computer use tasks Sonnet wins 83.9 to 77.6.66% similarUnverifiedGemini 4 Hydrogen scored 57% on the Terminal Bench 4.1 test, more than three times the previous generation.64% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1171959API
curl https://kongchang.com/api/v1/knowledge/claims/1171959MCP
get_claim(id=1171959)