Finally, some sanity in the AI pricing circus. Google just dropped Gemini 2.5 Pro—its “most intelligent model ever”—into public preview, and shockingly, it doesn’t require venture capital funding just to run a few API calls. At $1.25 per million input tokens, it’s cheaper than Claude 3.7 Sonnet ($3M) and OpenAI’s GPT-4o ($2.50M).
The AI Price War We Deserve
Google’s move is a direct middle finger to the “pay-to-play” model of AI access. While Anthropic and OpenAI nickel-and-dime developers with “premium reasoning” surcharges, Gemini 2.5 Pro undercuts them without sacrificing performance. Want proof? Even skeptics admit it’s the “most useful reasoning model yet.” But let’s not pretend this is altruism. Google’s playing chess while others fumble with checkers—undercutting rivals to lock in devs before they realize they’re hooked.
Why This Actually Matters
- Claude 3.7 Sonnet ($3M inputs, $15M outputs) now looks like a luxury sedan with a broken AC.
- OpenAI’s o1 reasoning model ($15M inputs) just got exposed as highway robbery.
- DeepSeek’s R1 started this price war, but Google just dropped a nuke.
The best part? No artificial scarcity. Google boosted rate limits because—gasp—they actually want people to use their tech.
The Fine Print (Because There Always Is One)
Sure, it’s still in “public preview,” meaning Google could yank the rug later. And yes, Vertex AI users have to wait a bit longer. But for now? This is the first time in years an AI pricing update didn’t feel like a scam. Bottom line? If you’re still paying OpenAI’s premium tax, maybe it’s time to switch teams. 🚀💰