OpenAI has announced a limited preview release of GPT-5.6, its latest large language model. The company claims it surpasses the capabilities of Anthropic’s Claude Mythos.
This development, reported by WatcherGuru, follows high anticipation after the success of GPT-5.5. Prediction markets had previously assigned a high probability to a late June release.
The move intensifies the competitive landscape between the two AI firms, particularly in areas like cybersecurity where Claude models are applied. Market observers are now monitoring official communications and performance benchmarks on leaderboards such as LMArena to verify the claim.
The announcement has already influenced prediction market odds for the model's public release and its initial performance scores.