Two OpenAI announcements, published a day apart in late July 2026, shape the AI landscape SMEs will use this fall. On one hand, GPT-5.6 prices dropped sharply. On the other, OpenAI says its models now reach more than one billion active users and two million businesses. For an SME leader, this is not just a double press release: it signals that generative AI has become a mass-market tool, at a fast-falling cost of entry. Here is what matters, with figures and sources.
Key takeaways
- On July 30, 2026, OpenAI cut GPT-5.6 Luna API prices by 80 percent: $0.20 per million input tokens (down from $1) and $1.20 per million output tokens (down from $6) (source: OpenAI, CNBC).
- GPT-5.6 Terra dropped 20 percent: $2 per million input tokens and $12 per million output tokens, down from $2.50 and $15 (source: OpenAI). Pricing for Sol, the most capable model, is unchanged.
- On July 31, 2026, CFO Sarah Friar announced that OpenAI's models reach more than one billion active users and more than two million businesses worldwide (source: OpenAI, Axios).
- According to OpenAI, users active for six months send 50 percent more messages per day and use the AI for twice as many task types as when they started.
- For an SME, cheaper access makes AI-driven automation more affordable, but a lower per-token price does not guarantee a lower bill if usage grows at the same time.
What changed with the July 30, 2026 price cut
OpenAI attributes the price cut to efficiency gains across its models and compute infrastructure. It also comes amid intensifying price competition, with Chinese models such as DeepSeek V4 or Kimi K3 priced well below their American counterparts. The cut applies only to API access: ChatGPT subscriptions (Plus, Business) keep the same price, but consume fewer credits for the same amount of Luna or Terra usage.
| Model | Price before (input / output, per million tokens) | Price since 07/30/2026 | Change |
|---|---|---|---|
| GPT-5.6 Luna | $1 / $6 | $0.20 / $1.20 | -80% |
| GPT-5.6 Terra | $2.50 / $15 | $2 / $12 | -20% |
| GPT-5.6 Sol | $5 / $30 | $5 / $30 | Unchanged |
Source: OpenAI; API pricing confirmed by CNBC and InfoWorld.
OpenAI is also introducing a "Fast Mode" for the API, replacing the earlier priority processing reserved for the Sol model. For an SME, Luna is now worth reconsidering for sorting, summarizing, or first-line support tasks, where its cost previously looked disproportionate compared with alternatives like Gemini Flash.
One billion users, two million businesses
The day after the price cut, Sarah Friar detailed, in an internal note picked up by the press, the scale of adoption of OpenAI's tools. The figure does not clearly separate weekly and monthly users, but it confirms a trend already tracked by firms like Sensor Tower: ChatGPT has become a mass-market tool, comparable in volume to major digital platforms.
OpenAI also observes a pattern of gradual adoption inside companies: a team starts with an isolated use case (writing, customer support, market watch), then usage spreads to other departments once trust is established. This mirrors exactly what LUWAI sees among its SME clients: one successful use case opens the door to bigger projects.
What this actually means for your SME
Audit current usage
Reserve Luna for simple tasks
Set a spending cap
In practice, an 80 percent cut on Luna means a text-processing workload that cost 100 euros a month can now run around 20 euros, at the same volume. That is a real argument for automating repetitive tasks (lead sorting, support ticket classification, first-draft replies) that were previously too costly to run at scale.
The low-price trap
A lower per-token price does not necessarily lower the total bill. According to OpenAI, users active for six months send 50 percent more messages per day and use AI for twice as many task types as when they started. A price cut matched by an equivalent rise in usage can leave the bill unchanged, or even push it up.
Limits to keep in mind
The one-billion-user figure is not publicly audited. It comes from a statement by OpenAI's CFO, without a detailed counting method or a clear split between weekly and monthly usage. Treat it as an order of magnitude, not a certified statistic.
The price cut only applies to Luna and Terra. Sol, the highest-performing model, stays at the same price. For tasks that demand the best reasoning quality, costs have not fallen.
Price competition remains a multiplayer game. Gemini Flash, DeepSeek V4, and Kimi K3 already offer much lower prices on some segments. An SME should keep comparing offers regularly rather than relying on a single vendor.
FAQ
Is GPT-5.6 Luna really 80 percent cheaper for every SME?
The cut applies to the per-token price billed through the API: $0.20 per million input tokens and $1.20 per million output tokens, down from $1 and $6. Standard ChatGPT subscriptions (Plus, Business) keep the same price, but consume fewer credits for equivalent Luna usage (source: OpenAI, Axios).
How do I know if my company is part of the two million OpenAI mentions?
OpenAI has not published a detailed counting method or a list of companies. The figure comes from a statement by CFO Sarah Friar, best treated as an order of magnitude rather than an audited data point.
Do I need to switch models to benefit from this price cut?
No, if you already use the GPT-5.6 API: the cut applies automatically to calls to Luna and Terra. If you use a standard ChatGPT subscription, no action is needed either.
Will this price cut continue?
There is no guarantee. OpenAI cites infrastructure efficiency gains, against a backdrop of intensifying price competition from DeepSeek, Gemini Flash, and Kimi K3. SMEs should keep comparing prices regularly rather than assuming cuts will continue.
Going further
The GPT-5.6 price cut confirms a trend LUWAI has tracked for months: the cost of accessing high-performing AI keeps falling, changing the return-on-investment math for many automation projects. Find our previous analyses of this trend on the Resources page. To see how French SMEs have already structured their AI adoption, check out our client success stories.
