Arena's October 19th snapshot ranked the October 17th model update No. 11 overall, based on 7,498 agent sessions.
By [Ryan Merket](https://runtimewire.com/author/ryan-merket)
· Published
Primary source: [X](https://x.com/arena/status/2112247114944360666)
Why it matters #
Arena's result puts a concrete task-cost figure beside an open-weight model's agent performance, giving developers a reason to test whether DeepSeek's efficiency gains hold in their own workflows.
DeepSeek's October 17th update to V4.1-Flash ranked No. 1 among open models and No. 11 overall on Arena.ai's Agent Arena leaderboard, with a reported median cost of $0.20 per task. Arena shared the result in a post on X, citing 7,498 sessions and a 5.13% net-improvement score.…
When you purchase through links in our articles, we may earn a small commission. This doesn’t affect our editorial independence.
source & further reading
runtimewire.com — original article
Artificial Analysis benchmarks how 36 AI models communicate
Cloudflare gives structured data tasks their own AI model with Clairic
Dream Labs raises $90M to sell robot control through a $3-an-hour API