Decisions API
API for decision models that answer questions about text, JSON or images with probabilities (yes/no, one of your options, or a rubric level) instead of text, billed per million input tokens.
Find alternatives to Decisions API
What it does
- Model inference
Sources
- Pricing
-
"The Decisions API is billed per million input tokens: $0.02 for pplx-decider-v1.2 and $0.01 for pplx-decider-flash-v1.2. Output tokens are free."
- Description
-
Official documentation · 18 Oct 2026
"The Decisions API answers questions about your content with probabilities instead of text: yes or no, one of your options, or a level on your rubric."
- What it does
-
Official documentation · 18 Oct 2026
Question types "noul", "choice" and "score"; models "pplx-decider-v1.2" and "pplx-decider-flash-v1.2".
- Sold within
-
Official documentation · 18 Oct 2026
"Any Perplexity API key works; create one at console.perplexity.ai if you need one."
- Name
-
Official documentation · 18 Oct 2026
"Decisions API".
- URL
-
Official documentation · 18 Oct 2026
The API's own quickstart page in the Perplexity API documentation, listed in the docs index as "Decisions API".
- Deployment
-
Official documentation · 18 Oct 2026
Hosted API: "a single JSON endpoint".
- Interfaces
-
Official documentation · 18 Oct 2026
"The Decisions API is a single JSON endpoint, so any HTTP client works."
Also from Perplexity
-
Perplexity Application
Search engine that answers questions in prose with numbered citations to the pages it drew on.
-
Comet AI agent
Web browser with Perplexity's assistant built in, able to read and act on the pages the user has open.
-
Agent API Model API
Model-agnostic API for search and agentic workflows — orchestrates across supported frontier models with built-in web search, URL fetching and reasoning controls; billed per request and per token.
-
Perplexity Computer AI agent
A general-purpose agentic 'digital worker' that orchestrates Perplexity's frontier-model lineup across tasks and tools to research, build, test, deploy and connect to 400+ apps, continuing work after the user steps away.
-
Perplexity Search API API service
Web search API that returns real-time, ranked web results as structured data with controls for domains, languages, regions and extracted page content, for feeding agents and applications.
-
Perplexity Embeddings API Model API
Embeddings API that generates standard and document-contextualized text embeddings (pplx-embed models, 1024 and 2560 dimensions, 32K context) for semantic search and retrieval, billed per million tokens.
-
Perplexity Orchestrator API Model API
Single API and key for open-weight models hosted by Perplexity, callable through the OpenAI Chat Completions, OpenAI Responses or Anthropic Messages formats, with automatic routing to a healthy deployment; in private preview.
-
Deep Research Application
Research mode inside Perplexity Computer that turns a complex question into a research plan, finds primary sources across many sites with every claim cited, and delivers reports, decks and dashboards from the user's files and the live web.
-
Portable Computer Application
Local-first edition of Perplexity Computer that runs the agent harness, planner and a 27B model on supported Windows and Linux PCs, with connected apps and local files, and asks before routing a task to one of 15+ cloud models.
Open Perplexity in the directory
Something wrong here? Send a correction — quote this product id: perplexity-decisions-api.