GPT-5.6, unlimited free chats and 1 billion users: what OpenAI's push means

For most of 2024, using ChatGPT for free felt a bit like being handed the keys to a nice car — but only for 20 minutes at a time. You'd hit a message cap, get bumped to a weaker model, or be told to come back tomorrow. That friction quietly nudged serious users toward the US$20-a-month Plus plan, or across the street to Google's Gemini and Anthropic's Claude.

OpenAI has just made that dynamic a lot more interesting. With the rollout of GPT-5.6 upgrades across both free and paid tiers, unlimited text chats for free users, and an expanded Codex coding agent that works from almost anywhere, the company is trying to do two things at once: hold the top of the market and squeeze the bottom so hard that competitors struggle to breathe.

Here's what the changes actually mean, and why they matter beyond the usual tech-news churn.

The free tier just got serious

The headline change for casual users is simple: free ChatGPT accounts now get unlimited text chats, according to Ubergizmo's reporting on the update. That's a significant reversal of the message-cap approach that has defined the free experience for years.

Unlimited text does not mean unlimited everything — image generation, advanced voice, deep research and the newest reasoning modes still typically sit behind quotas or a paywall. But for the enormous cohort of people who use ChatGPT as a writing assistant, a homework helper, a translator or a first-draft brainstorming partner, the practical experience is now closer to "just use it" than "ration it."

Combined with the GPT-5.6 upgrades that Firstpost reports are being rolled out across both free and paid ChatGPT plans, the free tier is no longer a demo. It's a genuinely capable product.

Why OpenAI is giving away more

On the surface, unlimited free access looks like a gift. It's not. It's a strategic move driven by scale, competition and the economics of AI distribution.

Seeking Alpha notes that ChatGPT has now crossed one billion users. At that size, OpenAI is not really competing with other chatbots anymore — it's competing with habits. The question is no longer "which AI is best?" but "which AI do you open without thinking?" Google has that reflex baked into search and Android. Microsoft has it baked into Windows and Office. For OpenAI, the only way to build that muscle memory is to remove every reason not to use ChatGPT.

Free-tier friction was one of those reasons. Removing it is cheap insurance against users drifting to Gemini, Claude, Meta AI or DeepSeek — especially in markets like Australia, where consumers are famously price-sensitive about subscriptions and where local businesses are still cautiously evaluating which AI stack to standardise on.

GPT-5.6 and the naming problem

The GPT-5.6 upgrade itself is more interesting than a version number suggests. According to Seeking Alpha, OpenAI has upgraded its GPT-5.6 Luna model and expanded Luna access as part of the same push. Firstpost frames the rollout as a broad refresh across free and paid plans rather than a single dramatic launch.

This is part of a pattern. OpenAI has increasingly moved away from the neat "GPT-4, GPT-5" ladder toward a fleet of specialised models — reasoning models, fast models, coding models, models tuned for particular tones or tasks — often given codenames rather than clean numbers. It's confusing for consumers, but it reflects the underlying reality: there is no longer one "best" model, only the best model for a given job.

For everyday users, the practical takeaway is that ChatGPT should feel smarter, more consistent and less prone to the odd hallucination in mid-tier queries — the sort of small quality bumps that don't make headlines but do decide whether someone keeps using a tool.

Codex from anywhere: the developer play

The other piece of the announcement is aimed squarely at software developers. OpenAI's update to Codex — its coding agent — now lets developers work with it from a broader range of environments and devices, rather than being tethered to a specific IDE or web interface.

This matters more than it sounds. Coding assistants are the single most commercially valuable application of large language models so far. GitHub Copilot, Cursor, Anthropic's Claude Code and a handful of well-funded startups have turned developer tools into the AI industry's most reliable revenue stream. If OpenAI can make Codex the default agent that follows a developer from laptop to phone to terminal to browser, it locks in a demographic that pays, evangelises, and builds on top of the platform.

For Australian tech workers — and the growing number of small dev shops and solo builders working remotely from Byron Bay, Hobart or the outer suburbs of any capital city — a coding agent that is genuinely portable is a bigger deal than another benchmark score.

What it means for the competitive landscape

Put the pieces together and a strategy comes into focus:

The obvious risk is cost. Serving unlimited text to a billion users is not free, even with efficiency gains in newer models. OpenAI is betting that a fatter free funnel will pay for itself through Plus, Team and Enterprise conversions — and through the sheer strategic value of being the AI that people default to.

The less obvious risk is regulatory. As ChatGPT becomes genuinely ubiquitous, expect regulators in the EU, the UK and Australia to look harder at questions of market power, training data, and the effect on local publishers and creative industries. "Free and unlimited" is a great consumer pitch. It is also exactly the sort of positioning that attracts competition inquiries.

What Australian users should actually do

If you've been paying for ChatGPT Plus mainly because you kept hitting message caps, it's worth re-testing the free tier for a fortnight. For writing, summarising, translating and everyday research, it may now be enough.

If you use ChatGPT for coding, image generation, long-context document work, or agentic tasks, the paid tiers still offer meaningfully more — and Codex's expanded availability is a genuine reason to look at OpenAI's developer tools if you haven't recently.

And if you're a business weighing which AI to standardise on, the calculus has shifted. The gap between free and paid is narrowing at OpenAI, but the gap between OpenAI's ecosystem and its rivals — in reach, in tooling, in developer mindshare — is arguably widening. That's not necessarily a reason to pick ChatGPT. But it's a reason to make the decision deliberately rather than by default.

The AI race in 2024 was about who had the smartest model. In 2025, it's turning into a fight about who has the stickiest habit. OpenAI has just made its move.

Related on Bleen

Sources