Treat High-Cost AI Models as Strategic Advisors to Cut Enterprise Spending
AI consultants advise firms to reserve expensive frontier models for planning tasks and delegate routine work to cheaper models, aiming to improve ROI on AI investments.
Enterprises seeking better returns on AI budgets are being urged to treat premium frontier models as advisory tools rather than workhorses. Ameya Kanitkar, CTO of Larridin, recommends using models such as Anthropic's Fable 5 to design workflow roadmaps, then assigning sub-tasks to less expensive models. Michael Murphy of Adaptovate adds that deploying the most powerful models for simple tasks like transcription or brief generation wastes resources.
Both consultants liken the strategy to hiring an expensive lawyer only for complex matters. Coinbase CEO Brian Armstrong has publicly predicted that within 12-18 months, the majority of AI workloads will run on far cheaper models, reserving the most capable systems for breakthroughs and agent coordination. The shift follows a broader move away from unrestricted token-burning practices, with firms exploring model-routing startups like OpenRouter and Concentrate AI to monitor and optimize spend.
Why it matters
Optimizing AI model usage can significantly lower corporate costs while preserving performance for critical tasks.
In this story