OpenAI's GPT-5.6 Sol Launching on Cerebras at Up to 750 Tokens/Second

OpenAI·medium signal

OpenAI's flagship GPT-5.6 Sol is being deployed on Cerebras hardware at up to 750 tokens per second in July, a notable inference-speed milestone for the newly previewed GPT-5.6 family (Sol/Terra/Luna). Pricing spans $5/$30 per 1M tokens for Sol down to $1/$6 for Luna. The rollout remains partly gated to 'trusted partners' at the U.S. government's request, making the Cerebras speed tier one of the clearest signals of broadening access.