Getting Started
- Visit chatgpt.com and sign up for a free account — you get GPT-4.1 mini access immediately.
- Upgrade to ChatGPT Plus ($20/mo) for GPT-4.1, o3, o4-mini, DALL-E, and higher usage limits.
- For Codex and the full model lineup, upgrade to ChatGPT Pro ($200/mo).
- For API access, go to platform.openai.com to get your API key.
Model Lineup
- GPT-4.1 — The latest fast multimodal model. Strong at coding, instruction following, and long-context tasks. Available in standard, mini, and nano variants.
- o3 — OpenAI’s most powerful reasoning model. Uses extended “thinking” for complex math, science, coding, and multi-step problems.
- o4-mini — Fast, efficient reasoning model. Great balance of intelligence and speed for everyday tasks.
- Codex (codex-1) — OpenAI’s autonomous coding agent. Runs in a cloud sandbox, reads your full repo, writes and edits code, runs tests, and creates PRs on GitHub. Works asynchronously — assign tasks and come back to finished work. Based on o3, fine-tuned with RLHF for software engineering.
- GPT-4o — Previous generation multimodal model, still widely used and available.
Key Features
- Codex coding agent — Connect your GitHub repo and assign coding tasks in natural language. Codex clones your codebase into a sandboxed environment, writes code, runs your test suite, iterates on failures, and opens a PR when done. Multiple tasks run in parallel.
- Deep reasoning with o3 — Extended thinking mode that works through complex problems step-by-step before responding, dramatically improving accuracy on hard tasks.
- Multimodal I/O — Text, images, audio, video, and file uploads in a unified chat interface. Generate images with DALL-E 3 inline.
- Real-time web search — Searches the internet for current information, news, prices, and events when needed.
- Code Interpreter — Executes Python in a sandboxed environment for data analysis, charting, file conversion, and computation.
- Canvas — A side-by-side editing workspace for writing and code, where you can collaborate with ChatGPT on documents and programs.
- Custom GPTs — Create and share tailored AI assistants with specific instructions, knowledge files, and API actions. Browse the GPT Store for community-built specialists.
- Memory — Remembers your preferences, projects, and context across conversations for more personalized responses.
- Voice mode — Real-time spoken conversations with natural-sounding voice synthesis and understanding.
- Operator and deep research — Agentic features that browse the web, perform multi-step research, and complete tasks autonomously.
// related tools
Claude
AI / Foundation Models & LLMs
Anthropic's frontier AI family — Opus 4.6, Sonnet 4.6, and Haiku for every use case
freemium
web
DeepSeek
AI / Foundation Models & LLMs
Open-source frontier models at a fraction of the cost — R1 reasoning shook the industry
oss
web git
Gemini
AI / Foundation Models & LLMs
Google's multimodal AI — 2.5 Pro with 1M+ context, Deep Research, and Google ecosystem integration
freemium
web