signalAI热榜2026-09-24
Artificial Analysis: Claude Opus 5.5 Tops Coding Agent Index, but Per-Task Cost Rises to $13.04
Claude Opus 5.5 scored 66 on the Artificial Analysis Coding Agent Index, beating Opus 5's 60 and Claude Fable 5.1's 62. Despite Anthropic cutting input/output token prices to $4/$20, per-task cost rose to $13.04, up 21% from Opus 5's $10.79, due to about 15.6 million tokens per task.
- for who
- Developers and teams using AI coding agents who prioritize benchmark performance
- why now
- New Opus 5.5 tops coding benchmarks, but costs $13.04 per task.
- what changes
- They now have a top-performing model that costs 21% more per task, forcing a trade-off between raw capability and budget
- to do
- Evaluate whether Opus 5.5's score gains justify the higher per-task cost in their own workflows
key points
- Scored 66, highest on the index, up from Opus 5's 60
- Per-task cost $13.04, 21% higher than Opus 5's $10.79
#coding agent#claude opus 5.5#cost analysis#benchmark
score
score 9 out of 10. 0-10: how dense the facts are, multiplied by how much you can do with them after reading. 8+ means the topic's evidence bar is met: benchmarks and availability for a new model, amount and investors for a funding round, revenue figures for a solo-money story. Below 5 an item does not enter the digest. A press release scores 3 or less, a reprint loses 2, anything older than 14 days loses 1, a headline that misleads loses 3.
read the source