OpenAI launched "ChatGPT Go" with paid features at ~20% the usual cost (help.openai.com)

OpenAI has launched ChatGPT Go, a lower-cost monthly subscription tier rolled out gradually in a set of mostly South and Southeast Asian countries (plus a few others) that packages expanded access to core ChatGPT features—most notably more usage of GPT‑5, image generation, file uploads, longer memory (bigger context window), and advanced data analysis tools (including a Python tool). Go also includes projects/tasks, custom GPT creation/editing, and works on WhatsApp and voice mode; billing is monthly, geo-restricted, and subscriptions can be managed or cancelled in-app. OpenAI positions Go as an affordable alternative (roughly ~20% of the usual cost), while still keeping higher-end capabilities like legacy models (e.g., 4o), connectors, agent mode, deep research tools and Sora video creation reserved for Plus/Pro plans. For the AI/ML community, Go is significant because it lowers the price barrier for experimentation with GPT‑5’s multimodal and reasoning capabilities, likely accelerating prototyping, data analysis workflows, and the creation of custom GPTs in emerging markets. Technical implications include broader real‑world usage of GPT‑5’s automatic “thinking” (reasoning) mode—users can force longer reasoning via “Try again → Think longer”—and increased demand on compute that may lead to dynamic usage limits. OpenAI will use conversations to improve models by default (with an opt‑out available), so wider adoption could also expand training/feedback data while leaving enterprise integrations (connectors) and advanced researcher features behind the higher tiers.