Anthropic launches cheaper, high-performance AI model Claude Opus 5
Anthropic has introduced Claude Opus 5, a new artificial intelligence model designed to offer performance comparable to its top-tier system but at half the cost. The San Francisco-based company, a significant competitor to OpenAI, claims this model excels in coding and knowledge-based tasks while utilizing fewer computing resources than previous iterations. This development directly addresses growing concerns among business clients regarding the high expenses associated with advanced AI systems. Claude's pricing, like that of competitors such as OpenAI's ChatGPT and Google's Gemini, is based on usage measured in tokens, with Opus 5 priced at $5 per million input tokens and $25 per million output tokens. This represents a 50% reduction compared to Anthropic's previous flagship model, Fable 5. Anthropic states that Opus 5 achieves performance levels close to Fable 5 across numerous applications despite its lower price point. The increasing scrutiny on company finances, particularly following Anthropic and OpenAI's June filings for stock market listings, has intensified focus on their operational efficiency and profitability. Regarding cybersecurity, Anthropic has deliberately excluded cyber-specific training for Opus 5, positioning it as less capable than its most advanced, restricted system, Mythos 5, in offensive cyber operations and potentially misused biological research. While Opus 5 can identify software vulnerabilities similarly to Mythos 5, it is significantly less adept at weaponizing these findings. Anthropic also notes that Opus 5's automated safety filters will engage approximately 85% less frequently than those on Fable 5, citing the reduced risk profile. Earlier, Fable 5 and Mythos 5 experienced temporary public access restrictions in June due to US Commerce Department export controls, which were subsequently lifted.
AI model pricing strategies are increasingly becoming a critical differentiator in a competitive market, especially as businesses evaluate the total cost of ownership for advanced AI solutions. Anthropic's introduction of Claude Opus 5 at a reduced price point, while maintaining high performance, suggests a strategic pivot towards broader market accessibility and adoption. This move could pressure competitors to re-evaluate their own pricing structures and resource optimization techniques. The company's deliberate approach to cybersecurity training in Opus 5, distinguishing its capabilities from more restricted models like Mythos 5, highlights the ongoing tension between AI advancement and responsible deployment. As AI systems become more integrated into critical infrastructure and sensitive research, the development of robust safety guardrails and transparent risk management frameworks will be paramount for sustained market trust and regulatory compliance.
AI-generated to prompt reflection — not editorial opinion, not advice, not a statement of fact. How this works.