AI firms prioritise cheaper, smarter systems over bigger models

The artificial intelligence race is entering a new phase as companies shift their focus from building the biggest models to developing cheaper, smarter systems that can choose the right AI model for each task.

Speaking to CNBC, Perplexity CEO Aravind Srinivas said the industry is moving beyond competing solely on model performance, with businesses now demanding better value, efficiency and flexibility from AI systems.

“The model alone is no longer the product,” Srinivas said. “It is the harness, the orchestration system that puts the model inside a very capable harness and pairs the model with a lot of tools.”

AI systems prioritise cost and efficiency

Rather than relying on a single powerful model, AI companies are increasingly building systems that automatically route tasks to the most suitable model based on complexity, cost and available data.

Routine customer service requests or internal workflows can run on cheaper open-source models, while more demanding tasks such as coding or advanced reasoning can be handled by premium models.

“The answer is always use whatever is the best for the task,” Srinivas said.

The trend comes as businesses become more focused on getting stronger returns from their AI investments, creating fresh challenges for companies such as OpenAI and Anthropic, which have built their businesses around premium frontier models.

Open models gain momentum

Perplexity recently previewed a new version of its computer-use system built around GLM 5.2, an open-weight model developed by China's Z.ai. The platform uses the lower-cost model for most tasks while switching to more advanced models only when necessary.

Benchmark general partner Peter Fenton believes open-weight models could dominate AI usage in the coming years.

“A maybe contrarian view that is becoming consensus is our belief that 90-plus percent of the tokens created will come out of open-weight models over the next 18 to 24 months,” Fenton said.

He added that the growing quality of open models could reduce the pricing power currently enjoyed by leading AI companies.

Enterprises seek greater control

The rise of open models is also encouraging businesses to run AI closer to their own infrastructure.

Ollama CEO Jeff Morgan said many enterprises prefer deploying smaller models locally, particularly in regulated industries such as healthcare, aviation and insurance.

“One thing is where the model’s from and where it was created and trained,” Morgan said. “But the more important thing to these businesses we speak to is where it runs and how it runs.”

According to Morgan, more than 85% of Fortune 500 companies now use Ollama to manage open AI models.

AI spending enters a new phase

The shift towards lower-cost AI systems could also reshape investment in cloud infrastructure, as some routine AI workloads move from large data centres to local devices.

Srinivas argued that affordable open-source models will help make AI more accessible to businesses.

“If you want the benefits of AI to be widely distributed to small businesses in America and American allied countries, then you really need AI to be a lot more affordable,” he said. “And open source is the only way to do that.”

As companies place greater emphasis on cost, efficiency and flexibility, the next stage of the AI race may be defined less by who builds the biggest model and more by who delivers the smartest and most economical AI systems.