STORY · MODELLER_
GPT-5.5 hallucinates three times more often than MIT-licensed GLM-5.2
An analysis shows that OpenAI's GPT-5.5 produces hallucinations three times more frequently than the openly available alternative GLM-5.2, which is released under the MIT license. The finding raises questions about performance differences between proprietary and open models despite their size.
WHY IT MATTERS
This challenges the assumption that larger models are automatically more reliable, and strengthens the case for investing in open models where reliability can be verified independently.
SOURCES
MACHINE-GENERATED SUMMARY This summary is written by machine from the sources below. We sort and explain — but we are a way into the field, not the final word. Check the source when something matters to you.