Shape the output - Perplexity

2026-02-04 · Source: perplexity.ai via Google News · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Software Development & Engineering, Cloud Computing & IT Infrastructure · Depth: Intermediate, short

Summary

Perplexity's API offers robust controls for managing output, addressing common needs beyond a single text block. Developers can enable streaming by setting "stream: true", receiving incremental Server-Sent Events with distinct types like "response.output_text.delta" for text chunks and "response.completed" for final usage data. For long-running tasks, "background: true" allows server-side execution and polling, supporting durable streams that can be reconnected using "GET /v1/responses/{id}?stream=true&starting_after=N". Response length is managed via "max_output_tokens" for a hard token cap and "text.verbosity" (low, medium, high) for soft preferences. Crucially, structured JSON output can be enforced using "response_format" with a defined "json_schema", ensuring machine-readable results and improving downstream system integration.

Key takeaway

For AI Engineers integrating the Perplexity API, understanding output control is crucial for optimizing application performance and user experience. You should implement streaming for interactive interfaces or long answers, and utilize background runs for computationally intensive tasks to maintain client responsiveness. Use structured JSON output with a defined schema to ensure seamless integration with downstream systems, avoiding brittle parsing and improving data reliability.

Key insights

Perplexity API offers granular control over response delivery and format, including streaming, background processing, and structured JSON.

Principles

• Stream responses for interactive UIs.
• Use background runs for long tasks.
• Define JSON schema for structured output.

Method

To control Perplexity API output, set "stream: true" for incremental events, "background: true" for server-side processing, or define "response_format" with a "json_schema" object. Manage length with "max_output_tokens" or "text.verbosity".

In practice

• Implement chat UIs with streaming.
• Run deep research tasks in background.
• Integrate model output directly into systems.

Topics

• Perplexity API
• Streaming API
• JSON Schema
• Background Processing
• Output Control
• Server-Sent Events

Best for: AI Engineer, Machine Learning Engineer, Software Engineer

Related on AIssential

• Sonar API - Perplexity — Perplexity's Sonar API provides web-grounded AI responses, offering OpenAI compatibility and native SDKs.
• Structured Outputs with LLMs: JSON Mode, Function Calling, and When to Use Each — LLM structured output methods vary in schema enforcement, suiting different application reliability and consistency needs.
• Building a Production ready RAG API — Production RAG APIs require robust backend engineering to handle ingestion, streaming, rate limits, memory, and caching at scale.
• Computer Usage Analytics - Perplexity — Perplexity offers an API for organizations to monitor detailed computer usage analytics across various datasets.
• Perplexity announces hybrid AI system that decides what runs locally or in the cloud — Perplexity's hybrid AI orchestrator intelligently routes tasks between local and cloud models for optimal privacy and efficiency.
• Structured outputs — Ollama now supports constraining LLM outputs to a JSON schema for reliable, consistent structured data.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by perplexity.ai via Google News.