[KongchangAI]
Unverified50% confidenceBenchmarkExact time

In actual testing on the POS reporting automation project, Gemini 3.1 Pro had a single-step latency of 28–45 seconds, consumed about 42% of quota, and failed after 3 refactoring iterations; Gemini 3.8 Flash High had a single-step latency of 3–6 seconds, consumed about 11% of quota, and succeeded on the first try.

1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/13/2026
First Seen
Valid until: 1/11/2027

Sources

Gemini 3.1 Pro Disappears from Antigravity: What Gemini 3.8 Flash Can and Cannot Do, Tested

Related Entities

Gemini 3.1 PromodelGemini 3.8 Flashmodel

Related Claims

UnverifiedAccording to a benchmark run headlessly via Antigravity on the same seven-constraint article planning task (three trials each, n=6, measured on 2026-09-27), Gemini 3.8 Flash (High) and Claude Opus 4.6 (Thinking) both scored 17/17 on machine-verified constraint adherence, but the median wall-clock time was 23 seconds for Gemini versus 67 seconds for Opus.74% similar

Cite This Claim

Stable URI
https://kongchang.com/claim/1013729
API
curl https://kongchang.com/api/v1/knowledge/claims/1013729
MCP
get_claim(id=1013729)