Anthropic Cuts Opus 5.5 Pricing 40% as OpenAI Counters With Cheaper GPT-6 Models

Anthropic released Claude Opus 5.5 on September 22, pricing it at $4 per million input tokens and $20 per million output tokens, a 40% cost reduction on typical workloads compared to Opus 5. The model matches Anthropic’s flagship Fable 5.1 on most tasks while running 30% faster. OpenAI responded 90 minutes later with GPT-6 Sol and Luna, cutting prices by half versus their predecessors, intensifying the race for cost-effective AI dominance.

Opus 5.5 matches top-tier performance at lower cost

Anthropic’s Opus 5.5 performs at the level of Claude Fable 5.1 on most work, according to the company’s announcement. Input tokens cost $4 per million, down from $5, and output tokens cost $20 per million, down from $25. Cache reads dropped 60% to 20 cents per million tokens. The model completed a 680,000-line code migration in under a day, work Anthropic said would have taken an engineering team weeks.

Safety testing and external validation

Opus 5.5 underwent testing by external evaluators including Frontier Design and METR before release. On Anthropic’s automated behavioral audit, the model scored highest of any system tested to date. It is more resistant to prompt injection than Opus 5 and less likely to take hard-to-reverse actions or act outside its boundaries. The model also improved clarity, with testers reporting 40% less verbose output without losing accuracy.

OpenAI launches cheaper Sol and Luna models

OpenAI released GPT-6 Sol and Luna on September 22, priced at half the cost of their GPT-5.6 predecessors. GPT-6 Sol costs $2 per million input tokens and $10 per million output tokens, down from $4 and $20. Luna costs 10 cents and 50 cents, down from 20 cents and $1.20. OpenAI attributed the price cuts to improvements in caching and inference efficiency. On the DeepSWE v1.1 software engineering benchmark, GPT-6 Sol matched Anthropic’s Fable at 68.8% while costing only 20% as much.

Benchmark competition and real-world performance

On Terminal-Bench 4.0, an agentic coding test, Opus 5.5 scored 66.4% versus 57.9% for GPT-6 Astra. On GDPval-AA v2.1, a knowledge work test, Opus 5.5 scored 1,846 versus 1,735 for Fable 5.1. Anthropic cautioned that benchmark margins have become less reliable at this level of capability. GitHub’s Chief Product Officer said Opus 5.5 solved more terminal tasks than Opus 5 in less than half the steps in VS Code.

Final Thoughts

Both companies are prioritizing cost efficiency over raw capability gains. With Opus 5.5 at 40% lower cost and GPT-6 Sol at half price, enterprise customers now have cheaper options for coding and knowledge work. The rapid release cycle signals intense competition for market share in AI services.

FAQs

Disclaimer:

The content shared by Meyka AI PTY LTD is solely for research and informational purposes.  Meyka is not a financial advisory service, and the information provided should not be considered investment or trading advice.