Oh boy! Where to start? New models, including 128k context are coming soon. GPT/OAI seems to have skipped by the 32k release all together and instead opted for 128k. Something we’ll be taking advantage of the moment it goes live! As an added benefit, the costs to these models has been significantly reduced. Costs have been halved or more depending on which OAI model is picked. Please see this blogpost for more details:
