OpenAI launched the GPT-5.6 series today! In our analysis, GPT-5.6 Sol delivers comparable performance to GPT-5.5 at a 13.2% lower cost.
This aligns with Sam Altman’s comment to CNBC that “every enterprise now is thinking about spend and the value they’re getting in exchange for AI.”
What does this mean for your organization? Achieving the right balance between performance and cost requires grounding your AI strategy in data, not vibes. The Findustry AI Benchmark includes hundreds of test scenarios and ensures our product leverages the optimal model for each task—if these insights can help with your AI transformation, feel free to contact our team.
Findustry AI CEO Jonathan Razi added: “It’s not your imagination: the pace of new model releases is getting faster and faster.
According to Stanford research, while just 9 foundation models were introduced in 2021, over 149 notable models were released in 2023. Now, this cadence is so rapid that the top model retains first place on leaderboards like Arena for an average of only 35 days.
To be a winner in applied AI, you have to be able to rigorously evaluate new models as soon as they are released. At Findustry AI, we run our Benchmark, which measures the performance of AI models for payments workflows, on the same day a new model comes out. The results directly inform how our product leverages base models and help our customers stay ahead in this fast-moving space.”