AI Agent Index
ByHeather MacAvelia·Independently reviewed·Last verified Jul 7, 2026

OpenAI's autonomous research agent that browses dozens of sources and produces cited reports in 5 to 30 minutes. GPT-5.6 models with MCP client connectivity. Free tier; Plus from $20/mo.

How we scored it

Autonomy

5/5

Integrations

4/5

Pricing clarity

5/5

Evidence

5/5

Setup

5/5

The facts

ChatGPT Deep Research is OpenAI's autonomous research agent, built into ChatGPT, that browses the web for 5 to 30 minutes to produce cited, comprehensive reports on complex topics. Rather than answering immediately, it plans a research strategy, browses dozens of sources autonomously, and synthesizes the findings into a structured report with citations. It runs on the GPT-5.6 model family (Sol, Sol Pro, Terra, and Luna), with GPT-5.5 Instant retained as the fast default, upgraded from the original GPT-5.2 launch. It scored 26.6% on Humanity's Last Exam and is state of the art on the GAIA benchmark for autonomous web research. Integration depth: MCP client connectivity, added February 2026, lets users connect Deep Research to external data sources and restrict searches to trusted sites, closing the public-web-only limitation of the original launch. By default it can use the public web, uploaded files, and connected apps with read-only access. Named connectors include document stores like Google Drive and SharePoint and authenticated industry data sources such as FactSet, PitchBook, and Scholar Gateway. Visual browser integration through agent mode, added July 2025, extends research to pages that need visual understanding. Users track progress in real time and can interrupt to refine the plan or adjust which sources it can access. Pricing follows ChatGPT plan tiers. Free includes a limited number of deep research queries. Go at $8 per month adds more messages and uploads but no expanded deep research. Plus at $20 per month includes expanded deep research with GPT-5.6 Sol, projects, tasks, and Custom GPTs. Pro from $100 per month provides the highest deep research limits with GPT-5.6 Sol Pro and priority access. Business at $20 per user per month billed annually ($25 monthly) adds admin controls, SAML SSO, and no training on business data. Enterprise is custom priced. Who it is not for: Deep Research is web-first, so it is not the best fit for structured literature review of indexed academic databases. Elicit ($49/mo billed annually), Consensus ($12/mo), and ResearchRabbit ($10/mo) handle systematic reviews of PubMed, arXiv, and Semantic Scholar better. Teams running high-volume research daily can also hit Plus query caps, which pushes them to Pro from $100 per month. Current state Q3 2026: the GPT-5.6 model family (Sol, Sol Pro, Terra, Luna) now powers Deep Research, with GPT-5.5 Instant as the fast default, upgraded from the original GPT-5.2 launch. MCP client connectivity is confirmed for connecting to external sources and restricting searches to trusted sites. Reports can be downloaded as Markdown, Word, or PDF. G2 shows 2,703 reviews at 4.6 for ChatGPT as of July 2026. SOC 2 Type 2, ISO/IEC 27001:2022, and ISO/IEC 42001:2023 are confirmed on OpenAI's security page, and OpenAI supports GDPR, CCPA, and HIPAA compliance through a Data Processing Addendum and Business Associate Agreement. Business and Enterprise tiers do not train on customer data.

Pricing

freemium · $20/mo

View pricing ↗

Segment

b2b

Setup

easy

Verified

Jul 7, 2026

Transparency

Public

Contract

Monthly or Annual

Data training

Opt-out

Autonomy

Autonomous

Capabilities

deep-researchweb-searchcitationsautonomousdata-analysisreportingmultilingual

Pros & Limitations

Editorial assessment

Pros

  • State-of-the-art autonomous web research benchmark scores (26.6% on Humanity's Last Exam, top of GAIA leaderboard) with GPT-5.6 models: meaningfully outperforms competing agents on complex multi-step queries requiring extensive source synthesis.
  • MCP client connectivity (added February 2026) lets users connect to external data sources and restrict searches to trusted sites: closes the public-web-only limitation of the original launch, enabling enterprise research grounded in authenticated industry sources.
  • Tightly integrated with ChatGPT Free, Go, Plus, Pro, Business, and Enterprise tiers: no separate purchase, deploys instantly to any team already on ChatGPT, eliminating procurement friction for organizations already using OpenAI products.

Limitations

  • Query caps constrain heavy users: Free is limited, Plus expands but still caps queries, and only Pro from $100 per month offers the highest limits for daily Deep Research use.
  • Confidence calibration is weak per OpenAI's own admission: the model often fails to convey uncertainty accurately, which matters for high-stakes research where false confidence is a worse outcome than admitted uncertainty.
  • It is not optimized for academic literature review against indexed paper databases: Elicit ($49/mo billed annually), Consensus ($12/mo), and ResearchRabbit ($10/mo) produce better systematic reviews of peer-reviewed sources than its web-first approach.

Technical Details

Deployment
webdesktopmobile
Model architectureGPT-5.5 Instant / GPT-5.6 Sol / GPT-5.6 Sol Pro
Avg setup timeunder 1 minute (existing ChatGPT account, select deep research in composer)
Autonomous rate5-30 minute autonomous multi-step web research with no human approval per step; produces fully-cited reports
Integrations
Google DriveSharePointFactSetPitchBookScholar GatewayMCP serversUploaded files
Security
SOC 2 Type IIISO 27001ISO 42001GDPRHIPAACCPA

Similar agents

Rating

4.8/ 5

Editorial score

How we score this →

Recognition

Rated 4.8 ★Category Leader · Q3 2026Transparent PricingListed 2026
Embed these on your site →

Industries

B2BEnterpriseSMBSaaSFinanceHealthcareLegalPharmaAgencies

Leave a review

Never displayed publicly.

Agent Stacks

See workflow stacks that feature ChatGPT Deep Research.

Compare

Related Content

Is this your tool?

Claim this listing free to get your embeddable badges, a review collection link, and the Verified checkmark.

Claim listing →

Free · Every Two Weeks

AI Agent Price & Rating Tracker

Price changes, new agent launches, acquisitions, and rating updates across 344+ AI agents. Verified against live vendor data, not vendor marketing.

No spam. Unsubscribe anytime. We never share your email.