Anthropic Launches Claude Opus 5 — Near-Fable 5 Performance at Half the Price

🚀 Anthropic Officially Launches Claude Opus 5

On July 24, 2026, Anthropic unveiled Claude Opus 5 — their new flagship AI model. Opus 5 is positioned as a much cheaper alternative to Fable 5 that can compete with it across several key benchmarks, particularly in coding and knowledge work domains.

The model maintains the same pricing as its predecessor Opus 4.8 at $5/MTok input and $25/MTok output. Compared to Fable 5's $10/$50 pricing, this is half the cost, making it attractive for businesses and developers scaling their AI usage.

📊 Agentic Coding — Opus 5 Takes the Lead

One of the most impressive results came from Frontier-Bench v0.1, where Opus 5 scored 43.3% on agentic terminal coding, significantly beating Fable 5 (33.7%) and GPT-5.6 Sol (34.4%). On the GDPval-AA v2 knowledge work benchmark, Opus 5 achieved 1,861 Elo, ahead of Fable 5's 1,747 and GPT-5.6 Sol's 1,736.

However, Opus 5 doesn't win everywhere. On DeepSWE v1.1, GPT-5.6 Sol leads with 72.7%, followed by Fable 5 (69.7%) and Opus 5 (68.8%).

🧩 ARC-AGI-3 — The Biggest Surprise

The ARC-AGI-3 benchmark, which measures novel problem-solving without memorized patterns, saw Opus 5 score 30.2% — nearly four times higher than GPT-5.6 Sol's 7.8%. Opus 4.8 managed just 1.5% on the same benchmark.

🛠️ Opus 5 Builds Its Own Tools

Anthropic describes Opus 5 as significantly better at checking its own work and improving through iteration. In one Frontier-Bench task, Opus 5 received a drawing of a machine part and had to create a 3D model in FreeCAD. The catch: the model had no way to view the drawing directly. Opus 5 wrote its own computer vision pipeline to extract geometry from raw pixels — no other model solved this task across five attempts.

⚡ Fast Mode and Effort Settings

Opus 5 features a 1 million-token context window. A new Fast Mode increases speed by 2.5x but doubles the price. Users can choose from five effort settings (low, medium, high, xhigh, max) that balance performance against token usage.

🔒 Cybersecurity — 85% Fewer Interventions

Opus 5's safety classifiers trigger about 85% less often than Fable 5's. It allows source code vulnerability research but blocks binary scanning, penetration testing, and exploit generation.

🔬 Scientific Research

Anthropic calls Opus 5 the most capable generally available model for scientific research, showing significant gains over Opus 4.8 in organic chemistry (plus 10.2 points) and protein-related tasks (plus 7.7 points).

🌍 Impact for Georgian Users

The release of Claude Opus 5 means Georgian developers and businesses get access to a more affordable yet powerful AI tool. For startups building on AI, it offers a significant cost advantage over Fable 5 while maintaining near-frontier performance.

💡 Conclusion

Claude Opus 5 represents Anthropic's strategic move to close the price-performance gap. It shows AI competition is no longer just about building smarter models — it's about making AI more accessible and efficient for everyone.

📖 Source