According to a recent LinkedIn post from Perplexity, the company is introducing “effort controls” within its Computer product’s model selector for Consumer Pro and Max users on the web. The post suggests these effort presets integrate orchestrator model selection and reasoning depth to calibrate how intensively the system tackles a task while managing computational efficiency.
The company’s LinkedIn post highlights that users may still manually adjust models, reasoning levels, and speed settings, indicating a layered control scheme on top of the new presets. The post also notes that Perplexity aims for “maximum intelligence at minimum cost” via multi-model orchestration, and plans to keep refining its model-agnostic harness and orchestrator or subagent configurations as frontier models evolve.
For investors, this feature rollout signals ongoing product differentiation in a crowded AI assistant market, emphasizing cost-performance optimization rather than a single-model strategy. The post suggests potential for improved unit economics if orchestration reduces expensive model usage while sustaining answer quality, which could support better gross margins on subscription tiers.
The availability of effort controls for paying Consumer Pro and Max subscribers indicates a focus on enhancing value for higher-tier customers and possibly improving retention among power users. As the feature expands to other platforms, Perplexity may deepen user engagement and usage frequency, which can be important leading indicators for future revenue growth and competitive positioning versus other AI productivity tools.