Hey everyone!
We have added gpt-oss-120b, OpenAI's newly released open-weights model.
This model is crazy good: o4-mini level performance at the API cost of gpt-4o-mini! It works crazy fast on Merlin, thinks deeply and works great with the power of agentic chat on Merlin.
According to our usage, this can reliably replace o4-mini/Gemini 2.5 Pro as your model for everyday tasks, owing to the lower cost.
Let us know how you like it!
Product Team @ Merlin