What ChatGPT Actually Does
ChatGPT handles an extraordinary range of tasks from one interface: writing, editing, summarising, translating, coding, data analysis, image generation (DALL·E), web search, document reading, and agentic tasks where it independently browses the web, fills forms, and completes multi-step workflows on your behalf. The 2026 version introduces two GPT-5.2 modes: Instant (fast, high-volume tasks) and Thinking (deep reasoning, complex research, code review), letting you dial up power as needed. Memory now carries context across all sessions — learning your tone, preferences, projects, and recurring tasks — while Projects lets you organise separate AI workspaces per client or goal. Voice has become a genuine first-class interface with real-time, natural speech-to-speech conversations, and Prism adds a dedicated long-form research workspace for collaborative deep dives.
Standout Pros of ChatGPT
ChatGPT's core superpower is versatility at world-class depth — it handles everything from a quick tweet rewrite to a full investment research report, switching contexts, tones, and formats effortlessly in a single conversation. GPT-5.2 Thinking mode is genuinely outstanding for complex reasoning, multi-step logic, and coding tasks, consistently outperforming most specialised tools. Memory and Projects make it feel less like a tool you prompt and more like a working relationship — it knows your business, your style, and your priorities without you re-explaining every session. The ecosystem depth (700+ integrations via Zapier, native Apple/iOS integration, custom GPTs, and Operator for agentic automation) means ChatGPT slots into virtually any workflow.
Weaknesses & Drawbacks
Despite massive improvements, ChatGPT still hallucinates — confidently presenting incorrect information as fact — which means anything high-stakes (legal, medical, financial) requires human verification. The model lineup has become confusing: GPT-5.2 Instant, Thinking, o3, o3-pro, and the new Go tier create decision fatigue for users who just want "the best one." Pricing has crept upward — Pro at $200/month is steep for individuals, and serious users quickly hit free-tier message limits. ChatGPT also trails Gemini on deep Android/smart home OS integration, and lacks multi-profile switching that teams sharing workflows genuinely need.
Pricing (2026)
Free
GPT-5.2 Instant, limited messages, lightweight Memory, image gen
$0
Go NEW
High-volume Instant access, affordable scaling tier
~$8–10/mo
Plus
GPT-5.2 Thinking, Deep Research, Prism, full Memory, DALL·E, Voice
$20/mo
Pro
Maximum limits, o3-pro, priority access, all features
$200/mo
Business / Enterprise
SSO, admin controls, no training on your data, SLA
Custom
Who It's For
Creators & Marketers
Scripts, captions, research, image gen, scheduling — all from one interface with Memory. Note: free tier limits apply, and specialist tools (Blort, 1of10) still outperform in narrow niches.
Developers & Technical Users
GPT-5.2 Thinking + code interpreter + API = most powerful coding assistant available. Gemini's deep OS integration edges it out for Android/device automation specifically.
Business Professionals
Projects, Prism, Voice, 700+ integrations, and Operator for agentic workflows make this a genuine productivity layer. $200/mo Pro is steep; Business tier requires negotiation.
Students & Casual Users
Free tier is genuinely capable; Memory makes it feel personal over time. Still hallucinates — requires critical thinking to use responsibly on anything factual.
