ChatGPT
OpenAI's conversational AI assistant used as a generalist model for research, drafting, reasoning, and ad hoc analysis across the team.
openai.com/chatgpt/teamAI ↗AI assistant for research, writing, strategy, and productivity.
2FA · destructive tool access revoked
Claude Team · Gemini Enterprise
OpenAI's conversational AI assistant, available as consumer and enterprise products, built on the GPT family of models.
Natural-language chat, reasoning, writing, coding (Codex), data analysis, image generation, deep research, agent mode, voice, file uploads, custom GPTs, projects, canvas, and record mode (meeting capture). Multiple model tiers.
Core product is the LLM itself; as of 2026 the default flagship is the GPT-5.x series. Synced apps/connectors cover SharePoint, Google Drive, Dropbox, Box, Outlook, Teams, Gmail, GitHub, Linear, and a growing set of partner MCP connectors (Atlassian Rovo, Stripe, HubSpot, Snowflake, Databricks, etc.). Custom MCP connectors supported on Enterprise/Business workspaces; "apps in ChatGPT" use MCP (Morningstar/PitchBook apps). Separate API billed per token.
Free; Go ~$8/month; Plus $20/month; Pro tiers ($100 and $200/month); Business $25/user/month annual ($30 monthly, 2-seat min); Enterprise custom (independent estimates ~$40–$60/user/month, historically ~150-seat minimum). API billed separately per million tokens. Pricing changes frequently.
60+ apps/connectors on Business/Enterprise; broad MCP support (both consuming external MCP servers and exposing apps); API and SDKs.
Business and Enterprise contractually exclude business data from model training. SOC 2 Type 2; aligned with CSA STAR; GDPR/CCPA support. Enterprise adds SSO/SCIM, RBAC, Enterprise Key Management (EKM), data residency, audit logs, HIPAA BAA on request. EKM disables synced-app connectors; Enterprise retains conversation data 30 days by default (configurable), not zero retention at query level.
Developed by OpenAI (San Francisco). ChatGPT launched November 2022; Enterprise launched August 2023.
Broadly useful for drafting, summarization, research synthesis, memo writing, and (via connectors/MCP) querying internal and licensed data (e.g., PitchBook/Morningstar apps in ChatGPT). Enterprise tier appropriate where data-training exclusion and admin controls are required.
Reliability for precise numerical/financial analysis is limited (hallucination risk); citation reliability lower without web/search; model retirements can disrupt workflows; EKM-vs-connector tradeoff; API costs are separate from subscriptions.