Expired60% confidenceFactTime unknown
Gemini 3 Pro spent 9 minutes and consumed 62,000 tokens in a UI designer portfolio test, yet the results were still poor.
2
Sources
60%
Confidence
Medium-term (~90 days)
Relevance
6/1/2026
First Seen
Valid until: 8/30/2026(expired)
Sources
Frontend UI Generation Showdown: GLM 4.7 vs Claude Opus vs Gemini vs Minimax — Who Wins?
bilibiliGoldenSpiderAI
Related Entities
Related Claims
UnverifiedAfter the release of Gemini 3.1 Pro, many developers were disappointed with its performance, believing that its single-instance capability had dropped significantly.71% similarUnverifiedGemini 3 Pro consumed a total of 223,000 tokens across five rounds of testing, the highest among the four models.70% similarUnverifiedGemini 3 significantly outperforms comparable models on the SWE-bench benchmark in agentic scenarios.70% similarUnverifiedGemini 3.1 Pro High achieved a pass rate of 84.17% on the SWE-bench test, ranking first, with 307,000 lines of code generated, a bug density of 614 per million lines of code, and 210 security issues per million lines of code.69% similarUnverifiedGemini 3.1 Pro took a total of five rounds of dialogue to barely complete the task of migrating an authentication system across projects69% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/20498API
curl https://kongchang.com/api/v1/knowledge/claims/20498MCP
get_claim(id=20498)