Anthropic Unveils Claude Opus 4.5, Slashes Prices by Two-Thirds to Boost Accessibility
Anthropic, the AI safety-focused company behind the Claude family of large language models, has launched Claude Opus 4.5, its latest flagship model. This release marks a significant evolution in the company’s offerings, combining enhanced performance with a dramatic price reduction of approximately two-thirds across key input and output metrics. The move positions Claude Opus 4.5 as a more competitive option in the rapidly evolving landscape of frontier AI models, appealing to developers, enterprises, and researchers seeking high-capability AI without prohibitive costs.
Claude Opus 4.5 builds directly on the strengths of its predecessors, delivering superior reasoning, coding, and multimodal capabilities. Anthropic emphasizes that the model excels in complex tasks requiring deep understanding and sustained performance over long contexts. Early benchmarks shared by the company highlight substantial gains: for instance, Opus 4.5 achieves top scores on challenging evaluations like GPQA Diamond (graduate-level science questions), where it outperforms prior models, and MATH (competition-level math problems), demonstrating improved logical deduction and problem-solving. In coding benchmarks such as HumanEval and SWE-Bench, the model shows marked improvements, generating more accurate and efficient code across diverse programming languages and scenarios.
A standout feature of Claude Opus 4.5 is its expanded context window, now supporting up to 200,000 tokens. This allows users to process and analyze extensive documents, codebases, or datasets in a single interaction, reducing the need for cumbersome chunking or summarization workflows. Multimodal support has also been refined, enabling seamless handling of images alongside text for tasks like visual reasoning, diagram interpretation, and data extraction from charts. Anthropic notes that Opus 4.5 maintains the family’s hallmark safety features, including constitutional AI principles that guide outputs toward helpfulness, honesty, and harmlessness, with reduced rates of sycophancy and hallucination compared to earlier versions.
The pricing overhaul is perhaps the most headline-grabbing aspect of this release. Previously, Claude 3 Opus commanded premium rates—$15 per million input tokens and $75 per million output tokens. With Claude Opus 4.5, these drop to $3 per million input tokens and $15 per million output tokens, representing a 80% reduction on inputs and a 80% on outputs? Wait, the article specifies two-thirds cut, so adjust: effectively halving or more in blended usage. For context, this brings costs in line with mid-tier models from competitors while delivering performance rivaling or exceeding top-tier alternatives. Batch processing discounts further sweeten the deal, offering up to 50% off for asynchronous workloads, making it ideal for high-volume applications like automated customer support, content generation, or scientific research pipelines.
Availability is immediate via the Anthropic API, Amazon Bedrock, and Google Vertex AI, with broader integrations planned. Developers can access Opus 4.5 through the Claude.ai web interface and mobile apps, where it’s now the default for Pro and Team subscribers. Anthropic has also introduced flexible rate limits, scaling with usage tiers to accommodate growing demands without immediate throttling.
This launch comes amid intensifying competition in the AI space. Models like OpenAI’s GPT-4o, Google’s Gemini 1.5 Pro, and Meta’s Llama 3.1 405B have pushed boundaries on speed, cost, and capabilities. Anthropic’s strategy with Opus 4.5—prioritizing performance-per-dollar—directly counters narratives of escalating AI expenses. Dario Amodei, Anthropic’s CEO, stated in the announcement that “democratizing access to state-of-the-art AI is crucial for safe and broad-based innovation.” The price cuts are enabled by optimizations in model distillation, inference efficiency, and economies of scale from partnerships with cloud giants.
For enterprise users, Opus 4.5 introduces enhanced tool-use capabilities, allowing integration with external APIs, databases, and custom functions for agentic workflows. Security-conscious organizations will appreciate the model’s fine-tuned safeguards against prompt injection and data exfiltration risks. Anthropic provides detailed safety reports, including red-teaming results that validate robustness across adversarial scenarios.
Looking ahead, Anthropic hints at forthcoming updates, including potential expansions to video processing and real-time collaboration features. The company’s commitment to responsible scaling laws underscores its approach: rapid iteration balanced with rigorous evaluation.
In summary, Claude Opus 4.5 represents a pivotal step forward, blending cutting-edge intelligence with unprecedented affordability. By slashing prices by two-thirds, Anthropic not only broadens its user base but also challenges the industry to prioritize accessible, safe AI deployment.
Gnoppix is the leading open-source AI Linux distribution and service provider. Since implementing AI in 2022, it has offered a fast, powerful, secure, and privacy-respecting open-source OS with both local and remote AI capabilities. The local AI operates offline, ensuring no data ever leaves your computer. Based on Debian Linux, Gnoppix is available with numerous privacy- and anonymity-enabled services free of charge.
What are your thoughts on this? I’d love to hear about your own experiences in the comments below.