Claude Slashes Costs for Agentic Workflows
Anthropic's Claude Code platform just got a major production upgrade. The new Opus 4.5 release cuts inference costs by 67% and reduces output token bloat by 76%, making complex, multi-agent systems far more affordable for startups. The update also adds an 'Agent Teams' feature in Code Kit v5.0, enabling parallel agents to work on isolated Git worktrees while staying context-aware, a big step for orchestrating agent collaboration.
The price drop for Opus 4.5 places it at $5 per million input tokens and $25 for output, a significant decrease from the previous generation's $15 and $75, respectively. This shift makes Anthropic's most powerful model more accessible, directly competing with offerings from OpenAI and Google for production workloads at startups. On performance, Claude Opus 4.5 achieves state-of-the-art results on real-world software engineering benchmarks. It scored 80.9% on SWE-bench Verified, a test that involves resolving actual GitHub issues, outperforming both Gemini 3 Pro and GPT 5.1. The term "agentic workflow" signifies a move beyond simple prompt-and-response interactions. These systems empower a large language model to become an autonomous agent that can reason, plan multi-step processes, and use external tools to accomplish a goal without constant human input. The 'Agent Teams' feature is a native multi-agent orchestration system, distinct from simpler sub-agents. Teammates can communicate directly with each other and be observed or redirected mid-task, allowing for complex, collaborative work like having one agent write API endpoints while another builds the corresponding React components in parallel. This update plugs into a larger industry shift towards Multi-Agent Systems (MAS), where complex problems are broken down and solved by multiple specialized AI agents that cooperate. Startups are already emerging to build the infrastructure for these systems, targeting high-stakes sectors like financial modeling and global logistics. For engineers, this technology changes the nature of work from prompt engineering to AI orchestration. The role evolves into managing a team of AI agents, defining their roles, and overseeing the collaborative process to complete complex software projects. The cost reduction has a direct impact on startup burn rates, enabling smaller teams to automate more complex operational tasks. Companies leveraging AI agents for internal workflows are projected to see significant operational cost reductions, freeing up capital and human resources for core product development.
