GPT-5.6 Sol Matches Claude Fable 5 on Code Arena — For 40% Less

GPT-5.6 Sol nearly matches Fable 5 on aggregated benchmarks at one-third the cost

OpenAI's GPT-5.6 Sol scores 59 points on the Artificial Analysis Intelligence Index, just one point behind Claude Fable 5. At $1.04 per task, it costs a third of what Anthropic's top model charges. In agentic coding,…