Perplexity launched Computer for Counsel. It is an agentic AI system built for legal teams. The product extends Perplexity Computer, the company’s LLM-agnostic agentic system. It is available now to Perplexity Enterprise and Max subscribers.
Lawyers lose hours to administrative work. Computer for Counsel targets that work directly. Nearly 75% of lawyers call administrative tasks a major time challenge, a Thomson Reuters survey found. The story is mostly architectural. It is an orchestration layer wired into the tools lawyers already use.
It is not a new legal research database. Perplexity is explicitly not trying to replace Westlaw, LexisNexis, or Bloomberg Law. Instead, it sits as a research, drafting, and workflow layer. That layer reasons over the open web, firm systems, and specialized legal sources.
The mechanics are agentic. The system decomposes a legal task into subtasks. It routes each subtask to a model and a data source. It then assembles the results into a brief, memo, or deal summary. Every output links back to its source. Attorneys verify a citation in seconds before it enters client work. Judgment and strategy stay with the lawyer.
Computer is powered by 20+ frontier AI models. It selects the best model for each subtask automatically. Research, reasoning, and contract work can each use a different model. Perplexity keeps the model pool current through ongoing evaluation. For legal teams, this removes the pressure to bet on one AI vendor.
Connectors run on the Model Context Protocol (MCP). MCP is an open standard for linking AI systems to external tools and data. Administrators can also install custom MCP connectors for internal systems.
Premium legal sources ground the answers. The connector list spans research, contracts, and document management.
| Source / Connector | Type | What it provides | Access at launch |
|---|---|---|---|
| Midpage | Legal research | US case law (federal + state appellate), statutes, regulations, a citator to check if a case is still good law | Uncapped for all Computer users; activate with @midpage |
| Deel | Compliance data | Worker classification, EOR rules, immigration, cross-border payroll across 150+ countries | Free, limited |
| LegalZoom | Contract templates | Customer agreements, employment contracts, NDAs via a template flow | Limited, coming soon, exclusive to Perplexity |
| Docusign | Contracts / e-signature | Agreement history and automated contract workflows | Available |
| NetDocuments / Box | Document management | Secure file systems and a legal context graph | Available |
| DeepJudge | Institutional intelligence | Grounds outputs in a firm’s prior work and accepted positions | Available |
| Clio (Vincent) | Legal research | Cited answers across 1B+ legal sources in 100+ jurisdictions | Coming soon |
| Carta / Ironclad | Equity / contracting | Cap tables, 409A data; AI contract repository search | Carta available; Ironclad coming soon |
App Connectors also reach Microsoft 365, Google Workspace, and 400+ other tools. Inside Microsoft 365, Computer drafts in Word and retrieves files from SharePoint. It references context from Outlook or Teams conversations.
Three current workflows show the agentic pattern in practice:
Computer for Counsel ships as a product, not an SDK. But it builds on Perplexity’s cited search. That primitive is exposed publicly through the Sonar API. The API is OpenAI-compatible and returns sources with every answer. Domain filters let you restrict grounding to trusted sites, the way a lawyer would.
# Perplexity's cited search primitive — the foundation Computer for Counsel builds on.
# This is the public Sonar API, not the Computer for Counsel product itself.
import os, requests
resp = requests.post(
"https://api.perplexity.ai/chat/completions",
headers={"Authorization": f"Bearer {os.environ['PERPLEXITY_API_KEY']}"},
json={
"model": "sonar-pro",
"messages": [
{"role": "user",
"content": "Is a non-compete signed in California enforceable in 2026?"}
],
# Restrict grounding to trusted sources, the way a lawyer would.
"search_domain_filter": ["law.cornell.edu", "courtlistener.com", "ca.gov"],
},
timeout=60,
)
data = resp.json()
print(data["choices"][0]["message"]["content"]) # the answer
for url in data.get("citations", []): # every source, for verification
print("source:", url)The pattern maps to the product. The model returns a grounded answer. The citations field returns the sources behind it. Computer for Counsel adds model routing, MCP connectors, and a review step on top. The lawyer still checks each cited claim before it ships.
Perplexity enters a crowded legal AI market. The positioning differs by design.
| Capability | Computer for Counsel | Westlaw / LexisNexis | Harvey | Microsoft 365 Copilot |
|---|---|---|---|---|
| Primary role | Workflow + research layer | Research database | Purpose-built legal platform | Productivity assistant |
| Model strategy | 20+ models, auto-routed | Proprietary stack | Multi-model, auto-routed | Microsoft + partner models |
| Good-law citator | Via Midpage | Native (KeyCite / Shepard’s) | Not its focus | No |
| Source-linked output | Yes, every answer | Yes | Yes | Partial |
| Reaches firm files | 400+ connectors via MCP | Limited | Connectors + storage | Microsoft 365 + connectors |
| Trains on your data | No | No | No | Per tenant settings |
Perplexity is not trying to out-Westlaw Westlaw. It targets the work before, around, and after formal research. Multi-model routing is no longer unique; Harvey routes across vendors too. The real differentiator is reach into the open web and everyday firm tools.
Check out the Technical details. Also, feel free to follow us on Twitter and don’t forget to join our 150k+ML SubReddit and Subscribe to our Newsletter. Wait! are you on telegram? now you can join us on telegram as well.
Need to partner with us for promoting your GitHub Repo OR Hugging Face Page OR Product Release OR Webinar etc.? Connect with us
Michal Sutter is a data science professional with a Master of Science in Data Science from the University of Padova. With a solid foundation in statistical analysis, machine learning, and data engineering, Michal excels at transforming complex datasets into actionable insights.
此内容由惯性聚合(RSS阅读器)自动聚合整理，仅供阅读参考。 原文来自 — 版权归原作者所有。