Every major ChatGPT update through July 2026: the GPT-5.5/5.6 transition, Dreaming V3 memory, Lockdown Mode, and what free vs paid gets you now.
The default model is now GPT-5.5 Instant, with GPT-5.6 rolling out in stages and GPT-5.4 retired on July 23, 2026. The two biggest product shifts of the year: Dreaming V3, a rebuilt automatic memory system, and Lockdown Mode, a hardened setting against prompt-injection โ both launched June 4. The desktop app was reorganized around a Chat / Work split with cloud-synced Work conversations.
This page tracks ChatGPT the product. For OpenAI's API, platform and company news, see OpenAI updates.
GPT-5.6 reached GA on July 9, 2026 after a limited preview from June 26, shipping as three explicitly named tiers:
| Model | Role | Input / Output per 1M |
|---|---|---|
| Sol | Flagship | $5 / $30 |
| Terra | Balanced, everyday work | $2.50 / $15 |
| Luna | Most cost-efficient | $1 / $6 |
Three weeks after GA, OpenAI cut Luna by 80% and Terra by 20%. An intra-quarter cut that steep on a fresh family is a competitive answer to Gemini 3.6 Flash at $1.50/$7.50 and Sonnet 5's $2/$10 introductory rate, not a promotion.
GPT-5 hid tier selection behind a real-time router in 2025 and users read it as silent downgrading. Published per-tier pricing is the direct consequence โ the clearest evidence that the 2025 launch lesson stuck.
The cadence lesson of the 5.x line: OpenAI now retires models on roughly a two-quarter rhythm โ build workflows against capabilities, not model names.
ChatGPT's automatic memory system was rebuilt from the ground up (launched June 4, 2026). Instead of the old fact-list approach, Dreaming V3 consolidates what it learns across conversations into evolving profiles that inform responses โ closer to "it knows how you work" than "it stored what you said." Memory remains user-controllable: view, edit, or wipe.
Also June 4: a hardened operating mode for high-risk contexts. With agentic features reading web pages, files and emails, prompt injection became ChatGPT's most consequential attack surface (the #1 AI security threat of 2026 industry-wide). Lockdown Mode constrains what external content can trigger, trading some capability for a materially smaller attack surface โ recommended for anyone running ChatGPT agents over untrusted inputs.
Memory is the stickiest feature ChatGPT has shipped and the hardest to leave behind. After a few weeks it demonstrably reduces re-explaining โ and it also means switching vendors now means rebuilding context you stopped noticing you had. Treat that as a real, if invisible, lock-in.
It's an admission that a chat window is the wrong container for delegated work. Work conversations sync, carry project context, and host the agent โ which is a workspace product wearing a chatbot's interface.
Free tier: entry-level model, rate-limited browsing, image generation and a monthly Thinking allowance โ genuinely usable, deliberately capped. Plus ($20/month): current flagship access, higher limits, memory and agent features. For a cross-vendor free-tier comparison, see best free AI tools.
Pay if you hit rate limits weekly, if memory matters to how you work, or if you need the current flagship rather than the entry model. Stay free if you use it a few times a week โ and note that Gemini's free bundle covers casual needs more generously.
Sources: OpenAI ChatGPT release notes ยท Releasebot ChatGPT tracker ยท AI Business Weekly on July features ยท SolidAITech 2026 guide
Last updated: July 29, 2026
Anthropic's 2026 in business terms: the $65B Series H at a $965B valuation, confidential IPO filing, $47B run-rate, and enterprise adoption of Claude.
Every major Claude update through July 2026: the Claude 5 family, Reflect dashboard, MCP upgrade, Claude Code milestones, pricing and access.
Copilot's 2026: Agent Mode general availability, agentic code review that ships fix PRs, the Copilot CLI, desktop app, and much faster indexing.
Every major Gemini update through July 2026: the 3.5/3.6 Flash generation, Gemini Omni multimodal creation, Nano Banana 2, and the agentic-era strategy.
Midjourney's 2026 in full: V8 alpha to the V8.2 default, native 2K without upscaling, 50% faster and 25% cheaper renders, and 21-second video.
Notion's 2026: autonomous Custom Agents, the credit billing switch on May 4, agents that read your files, 35โ50% cheaper runs, and what it costs now.
OpenAI's 2026 from the platform side: the GPT-5.x release-and-retire cadence, the Sora shutdown, API changes, and where the company stands mid-year.
Perplexity's browser year: Comet free on every platform since March, iPad and iOS upgrades, Deep Research generating decks, and enterprise controls.
