GPT-5.6 Sol tops the frontend leaderboard
Fresh metrics from Arena AI delivered an unexpected result: GPT-5.6 Sol xHigh shares first place with Claude Fable 5 in Code Arena: Frontend. This is the first time an OpenAI model has reached the top of the leaderboard.
Sol costs $5 per million input tokens and $30 per million output tokens, roughly half the price of Fable. Yet it can already compete with it in frontend development, web app creation, and agentic coding.
Anthropic now has a real competitor. Removing Fable from its subscription or restricting access has become risky, as users could simply switch to GPT-5.6 Sol. It looks like the development of Opus 5.0 may now need to accelerate.
Top comments (2)
The benchmarks are misleading. I have been working on a project for the last 8 days using Fable 5, and I have built a great UI along with amazing backend code. As this is an MVP, I have used Django (full-stack). Today, I decided to use GPT 5.6 Sol (Extra High) to improve the UI/UX and to improve the backend code where possible. I used the plan mode in Cursor. Sol (xHigh) took over 2 hours to complete all 9 to-dos it came up with, and at the end, the results were really disappointing. The UI/UX was messed up, and it didn't make any meaningful improvements to the backend code. I spent $274.10 USD on this refactor, and in the end, I had to hard reset to the last commit of the code assisted by Fable 5. Fable 5 is way ahead of GPT 5.6 Sol in real-world UI/UX and backend code, as far as I have understood.
Excellent. I saw the modal option in Cursor this morning. I'll have to give it a try.