The AI Price War Just Changed Shape: DeepSeek Is Raising Prices, ChatGPT Went Free, and Meta Wants Your Data
In the last seven days the AI market moved in three directions at once. DeepSeek said prices are going up, significantly. OpenAI made its newest model free with unlimited chats. And Meta released a model so cheap it is basically asking you to trade your data for a discount.
If you build websites, sell software, or run a business that leans on AI tools, this matters more than the headlines suggest. Here is what actually happened and what I think you should do about it.
Three moves in one week
1. DeepSeek: the cheap era is ending
On August 5 and 6, DeepSeek updated its official pricing page with a warning: it plans to raise overall API pricing in the near future, and the increase is expected to be significant. No number yet, no date. Just a heads-up for developers who built products on its famously low rates.
Current prices for DeepSeek V4-Flash are $0.14 per million input tokens and $0.28 per million output tokens. Those numbers made DeepSeek the default for cost-sensitive projects. If the increase is as large as the warning suggests, every app that depends on it needs a backup plan.
2. ChatGPT: free users get unlimited text chats
OpenAI announced that GPT-5.6 Luna becomes the default model for Free and Go users, with unlimited text chats. No rate limit on everyday conversations. A new Think button gives free users deeper reasoning for harder questions.
ChatGPT now has 1 billion weekly users. OpenAI is betting that giving away its latest model expands the market faster than charging for it. It also says factual errors dropped about 62% with Luna and 68% with the upgraded Sol model compared to the previous generation. Those numbers are self-reported, so I'd take them with a grain of salt, but the direction is consistent with what independent tests have been showing.
3. Meta: a model that pays you (in savings) for your data
Meta launched Muse Spark 1.2 with two price tiers. The standard tier costs $1.25 per million input tokens and $4.25 per million output. But there is a second tier, the "contributor" SKU, priced dramatically lower, around $0.10 and $0.20 respectively. The trade-off: your data may be used to improve Meta's models.
This is the most honest pricing model in AI so far. You are not paying with money. You are paying with data, and Meta tells you upfront. Whether that trade is worth it depends on what you build and who your customers are.
What this means for your website
If you use AI in your business, here is the practical translation:
- Do not build on one model. DeepSeek's warning is the reminder. The model that is cheapest today may not be tomorrow. Design your stack so you can switch providers without rewriting everything.
- Free tools change user expectations. If ChatGPT is free with unlimited chats, visitors expect your website's AI to feel instant and generous too. A slow, limited chatbot now looks worse by comparison.
- Data is the new currency. Meta's contributor tier is the clearest example yet. Every "free" or "discounted" AI offer is a trade: convenience for your data. Decide consciously what your business is willing to share.
What to do this week
- Audit your AI dependencies. List every tool and API your website uses. Note which ones could raise prices without warning.
- Keep a fallback model. Test a second provider now, while prices are low. Switching later, under pressure, is expensive.
- Check your chatbot quality. With free unlimited ChatGPT, the bar for "good AI" on your site just went up. Test yours with real visitor questions.
- Read the data terms. Before you accept any discount, know exactly what the provider can do with your customers' data.
The bottom line
Prices are going up, models are going free, and data is becoming the real currency. The businesses that win are the ones that stay flexible: no single vendor lock-in, a clear policy on data, and AI that actually serves visitors instead of just decorating a page.
Not sure how your website's AI holds up? Get a free audit at softgrama.com/audit. And if you want to understand where agents are taking this next, read about AI agents managing other AI agents.