Arena reports time and cost gaps among top models on Agent Arena long-horizon tasks

AIArena reports that Gemini 4 Carbon costs a median $10.51 per hour-long task on Agent Arena, versus $21.13 for Claude Opus 5.5 and $31.08 for GPT-6 Astra. Arena says Claude Opus 5.5 still scores higher but costs twice as much per hour of work. The comparison covers nearly 30,000 tasks run for over an hour in the last 14 days, averaging 332 tool calls per task.

AIArena says Google's Gemini 4 Argon and Xiaomi's MiMo-V2.6 Flash and Pro gain frontier spots on the most expensive 5% of Agent Arena sessions, where OpenAI and Anthropic hold 7 spots versus 10 of 11 on cheaper tasks. More than 1,000 sessions costing $100 or more occurred in the last two weeks.

AIArena's Agent Arena comparison of the top model from each of five leading labs shows a 55-minute gap in time spent on long-horizon tasks. Claude Opus 5.5, ranked #1 on Agent Arena, takes the longest at every task length, peaking at 1 hour 51 minutes. Kimi K3.1 Pro spends the least time at every length, with 56 minutes on long-horizon tasks.