Unverified50% confidenceEventExact time
In the large-scale code task "Dead Three Code" test, Gemini 4 Hydrogen ran up over 100 yuan in unpaid charges using the official API, the task took about four hours, and it finally finished with a score of 77.9 in the anti-gravity subscription environment, which the UP host considers not entirely fair.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
11/14/2026
First Seen
Valid until: 2/12/2027
Sources
Gemini 4 Hydrogen Hands-On Test: Has Google Finally Caught Up in Coding?
bilibili程序员阿江-Relakkes11/13/2026
Related Entities
Related Claims
UnverifiedMultiple users report that Gemini 4 Hydrogen's weekly quota is consumed faster than expected.71% similarUnverifiedGemini 4 Hydrogen ranked around 7th in the bridge-building experiment, but three problems have not yet been completed due to API return issues.71% similarUnverifiedOn Terminal Bench 4.1, Gemini 4 Hydrogen reduces fabricated statements by more than 5 times and handles failed tool calls more stably.70% similarUnverifiedHydrogen inherits the Gemini 4 safety stack, reducing overreach by up to 81% and under-service by up to 79% in agentic workflows, and reducing misaligned behavior and task shortcuts by up to 12x in tool-use and coding evaluations.69% similarUnverifiedAccording to benchmarks provided by Google, Gemini 4 Hydrogen was compared with Sonnet and Terra and won in most tests.69% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1171394API
curl https://kongchang.com/api/v1/knowledge/claims/1171394MCP
get_claim(id=1171394)