Skip to content
AI Primer
release

Perplexity adds Finance Search to Agent API with live data and FinSearchComp T1 lead

Perplexity added Finance Search to the Agent API with licensed real-time market data and cited web sources in one tool call. The company says it led FinSearchComp T1 on live-data accuracy and lowest cost per correct answer, so teams building finance agents should evaluate it against their current stack.

4 min read
Perplexity adds Finance Search to Agent API with live data and FinSearchComp T1 lead
Perplexity adds Finance Search to Agent API with live data and FinSearchComp T1 lead

TL;DR

  • Perplexity added Finance Search to the Agent API, and perplexity_ai's launch post says one tool call now returns licensed financial datasets, real-time market data, and cited web sources.
  • The feature is aimed at live finance questions, because perplexity_ai's follow-up lists prices, fundamentals, earnings, filings, and market context as first-class query types.
  • In Perplexity's own FinSearchComp T1 comparison, perplexity_ai's cost chart shows Finance Search at $0.0087 per correct answer, below GPT 5.5 with built-in web search at $0.0355 and Exa Answer at $0.0556.
  • AravSrinivas' thread adds a latency claim: Finance Search needed the fewest minutes after market close to return the right answer on the benchmark's live-data track.
  • The API launch lands a day after Perplexity's Professional Finance launch put licensed finance data and 35 analyst workflows into Perplexity Computer, extending the same vertical-data push into agent tool calls.

You can read the official Finance Search post, compare it with Perplexity Computer for Professional Finance, and note that the company also shipped premium health sources the same week. The interesting bit is not just another search endpoint. It is Perplexity turning licensed vertical data into a product surface for both end-user workflows and agent APIs.

Perplexity's packaging is unusually compact. According to perplexity_ai's launch post, the tool call bundles three things that usually live in separate integrations: licensed financial datasets, live market data, and cited web results.

perplexity_ai's follow-up narrows the initial scope to five query families:

  • prices
  • fundamentals
  • earnings
  • filings
  • market context

That lets the company pitch Finance Search as a retrieval primitive for valuation lookups, earnings recaps, and market monitoring without wiring up separate market-data vendors first, as described in the official blog post.

FinSearchComp T1

Perplexity anchored the launch to FinSearchComp T1, a live-finance benchmark the company says measures both accuracy and post-close freshness. In perplexity_ai's chart, Finance Search posts the lowest cost per correct answer in the cohort at $0.0087.

The same chart places the other compared systems at:

  • Claude Opus 4.7 with built-in web search: $0.3878
  • Exa Answer: $0.0556
  • GPT 5.5 with built-in web search: $0.0355
  • Gemini 3.1 Pro with built-in web search: $0.0333
  • Parallel Lite: $0.0313

The time-sensitive angle is separate from the cost chart. AravSrinivas' thread says Finance Search also needed the fewest minutes after market close to reach the right answer, while the line chart attached to the launch thread shows GPT 5.5 catching up later in the window.

Professional Finance

The API feature shipped one day after Perplexity's end-user finance product. In perplexity_ai's Professional Finance launch, the company said Perplexity Computer now accepts licensed data from Morningstar, PitchBook, Daloopa, and Carbon Arc, plus 35 dedicated workflows for recurring analyst work.

The launch thread broke those workflows into concrete outputs instead of generic research mode branding:

Finance Search looks like the API-side version of that same bet: licensed data plus citations as infrastructure, not just UI.

Premium sources

Finance was not a one-off vertical. A day earlier, perplexity_ai's health-sources launch said Perplexity and Computer added premium medical sources starting with NEJM and BMJ Group, with more journals and clinical databases on the way.

That makes the week's releases read as a broader sourcing strategy:

Across both launches, Perplexity kept repeating the same product claim. As AravSrinivas' post put it, every output number in the finance product is traceable back to a source.

Share on X