88Z AI, 23 September — Anthropic has released Claude Opus 5.5, the first model in its Claude 5.5 series, setting new state-of-the-art records in agent coding, computer use, and knowledge work. It scored 66.4% on Terminal-Bench 4.0, outperforming GPT-6 Astra by 8.5 percentage points, and achieved 1846 Elo on the GDPval benchmark. Additionally, the model is 40% cheaper, 30% faster, and features significantly improved writing and communication capabilities.
At a glance
- Claude Opus 5.5 sets new SOTA records on Terminal-Bench 4.0 and GDPval benchmarks, outperforming competitors.
- The new model is 40% cheaper and 30% faster, with significantly enhanced writing and communication skills.
The update
Anthropic has officially launched Claude Opus 5.5, the first flagship model of the Claude 5.5 series. This model comprehensively refreshes state-of-the-art performance standards in areas such as agent coding, computer operation, and knowledge work. In specific benchmarks, it achieved a score of 66.4% on Terminal-Bench 4.0, surpassing GPT-6 Astra by 8.5 percentage points. On the GDPval benchmark for knowledge work, it reached a high score of 1846 Elo. Beyond performance breakthroughs, Claude Opus 5.5 features significant optimizations in cost and efficiency. The overall price has been reduced by 40%, while operating speed has increased by 30%. Furthermore, the model's writing and communication capabilities have been notably enhanced, allowing it to handle various text interaction tasks more effectively.
Sources & editorial note
This article is an AI-assisted synthesis by 88Z AI based on the public sources below. Product claims and availability are attributed to their publishers. It does not represent independent testing or first-hand reporting.
- APPSOSource published
Source dates above refer to the original announcements; the date under the headline is this article’s publication on 88Z AI.