Claude Opus 5 Launched: Half the Price, 22% Better at Coding — Is This the New AI Default?

Why this matters

Anthropic released Claude Opus 5 today, and the developer community is taking notice. We pulled the official announcement and early-access reports from Cursor, Devin, and Zapier to verify the claims. Here's what 1603 Hacker News points worth of discussion actually means for teams using AI daily.

What we verified (real data)

Pulled from Anthropic's launch page on July 25, 2026, and cross-checked against the ARC-AGI leaderboard + Artificial Analysis:

• Frontier-Bench v0.1 — Opus 5 surpasses all other models
• ARC-AGI 3 — Opus 5 score is 3x the next-best model
• Zapier AutomationBench — 1.5x pass rate vs next-best
• OSWorld 2.0 — Beats every model at any cost
• CursorBench 3.2 — Within 0.5% of Fable 5 peak, at half the cost
• Internal coding evals — 22% improvement over Opus 4.7

Who is using Opus 5 already

• Devin (Cognition): Particular strength on difficult debugging and root-cause analysis tasks
• Cursor: Within 0.5% of Fable 5 with same behaviors at half the cost
• Zapier: Topped AutomationBench without spending more tokens. 100% pass rate
• Lovable: Up 22% over Opus 4.7, but more importantly — steadier
• Genomics teams: Behaves more like a careful scientist than any model we've run

Bottom line

Opus 5 delivers near-Fable 5 intelligence at Opus speed and cost. Same Pro price tier ($20/mo), stronger default model, fewer usage limits. The new state-of-the-art for coding, knowledge work, and agentic tasks.

**Disclosure**: This article was researched using Agent-Reach (open-source CLI by Panniantong) and cross-verified against Anthropic's launch page. We use Anthropic and partner affiliate links in this article — see our disclosure policy.