Unverified50% confidenceBenchmarkExact time
GPT-OSS-120b has nearly a 50% chance of making JSON formatting errors when used outside of BAML.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
9/22/2026
First Seen
Valid until: 12/21/2026
Sources
Related Entities
Related Claims
UnverifiedWhen using GPT-OSS-120B with Lemonade during the evaluation, the attempt failed after two tries of approximately 90 minutes each because Lemonade could not access the internet or read local files.62% similarUnverifiedThe perplexity of the GPT series of models on Wikipedia text has decreased from approximately 29 for GPT-2 to an even lower level for GPT-4.59% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/940128API
curl https://kongchang.com/api/v1/knowledge/claims/940128MCP
get_claim(id=940128)