Looks like o1 performance without reasoning. Pretty good but seems reasonable that they didn’t want to call this 5 as they’ve already got a product out there that is as performant.<p>Another notable thing here is a big drop in hallucination rate as measured by their benchmarks (for whatever those are worth).
Insane pricing - $75.00 / 1M input tokens & $150.00 / 1M output tokens. They mention it's a big model but it's hard to imagine inferencing costs being 20X higher than 4o.<p>I'm assuming they're doing everything possible to get everyone onto the reasoning models? This seems like it's going to be a short lived model.