We've launched a host of new models on Merlin.
GPT-4.1 Series
We've added three brand-new OpenAI models:
- GPT-4.1: Upgraded version of GPT-4o with better coding skills and understanding of complex instructions. Costs 15 ⚡ per non-Live response.
- GPT-4.1-mini: Similar quality to GPT-4o but faster and cheaper. Costs 3 ⚡ per non-Live response.
- GPT-4.1-nano: Replacement for GPT-4o-mini with a 100k token context window in Large Context mode. Costs 1 ⚡ per non-Live response.
Grok Models
Two new models from xAI are now available (finally):
- Grok 3: Excellent at math, reasoning, and coding problems. Costs 25 ⚡ per response.
- Grok 3 mini: Faster, cheaper version that's still great at math and coding. Costs 2 ⚡ per response.
We have plans to increase the context window on GPT-4.1 series models soon. Let's see how that goes!