[KongchangAI]
Unverified60% confidenceBenchmarkExact time

At the ultra-high thinking setting, Neon cost $6.44 per task to complete the full benchmark suite, 36% cheaper than Opus 5.5 Max's $10.04, and its time to first answer was six times faster (56.5 seconds vs 338.6 seconds); GPT 6.2 Sol cost $2.10 per task, but Neon's output speed of 110 tokens per second was 47% faster.

2
Sources
60%
Confidence
Short-term (~14 days)
Relevance
11/21/2026
First Seen
Valid until: 12/5/2026

Sources

Related Entities

Related Claims

Cite This Claim

Stable URI
https://kongchang.com/claim/1194490
API
curl https://kongchang.com/api/v1/knowledge/claims/1194490
MCP
get_claim(id=1194490)