Breaking

Claude Opus 5.5 Takes No. 1 Spot on Code Arena WebDev at 60% Lower Token Price Than GPT-6 Astra

September 24, 2026 at 09:14 EDT

Foundation Models
AI Agents
Software Dev & Coding

Claude Opus 5.5 has moved to the top of Code Arena's WebDev ranking, giving Anthropic a fresh coding-model win days after the model's launch. The leaderboard lists Claude Opus 5.5 Max at 1,818 points, ahead of GPT-6 Astra Max at 1,792, while charging $4 per million input tokens and $20 per million output tokens versus Astra's $10 and $50 pricing.

September 24, 2026 · WebDev ranking

Claude Opus 5.5 Takes the WebDev Lead

Claude Opus 5.5 Max reached 1,818 points, ahead of GPT-6 Astra Max, while its listed input and output token prices are 60% lower.

1,818

Top WebDev score

Point lead over GPT-6 Astra Max

60%

Lower input and output token prices

Token context window

Score gap versus earlier leaders

26 ptsLead over GPT-6 Astra Max

126 ptsGain over Claude Opus 5 Max

The bigger move is the jump from the prior Opus Max entry; the Astra lead is real on the table but narrower.

Why developers care

The result points to stronger web-app generation, refactoring and multi-step coding at lower listed token prices.

What still needs caution

The top two models have overlapping score intervals, and real production cost depends on effort level, output length and agent loops.

Selection lens

Leaderboard score→Token price→Effort setting→Completed task cost

Continue reading

The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.