Back
The DecoderHotThe Decoder2026-04-23

OpenAI Launches GPT-5.5 — Calls It a 'New Class of Intelligence' While Doubling API Prices

OpenAI unveiled GPT-5.5 today, framing it as a qualitative leap rather than incremental upgrade. The model arrives with API pricing roughly double that of previous versions, a bet that the capability jump justifies the cost — and a signal that the era of rapidly falling AI prices may be ending.

Original source

OpenAI today launched GPT-5.5, positioning it not as a routine version bump but as "a new class of intelligence" — language that signals the company believes it has crossed some qualitative threshold in the model's capabilities. The announcement landed on Hacker News with 608 points within hours, suggesting the developer community is paying close attention.

The most immediately controversial detail is pricing: API access to GPT-5.5 costs roughly double that of its predecessor. OpenAI appears to be betting that the capability gains justify the premium — a significant bet in a market where developers have grown accustomed to falling costs per token. If the jump is real, it's a defensible move. If it's incremental capability with an aggressive price tag, expect backlash.

Full benchmark details and capability specifics weren't immediately available at publication, but the positioning against "previous versions" suggests this isn't simply a GPT-5 rename. The timing is notable: it comes as Google unveiled 8th-gen TPUs and Kimi K2.6 dropped open weights at frontier performance — OpenAI is playing offense to maintain mindshare at the top of the capability stack.

For developers already managing AI cost at scale, a doubling of API pricing on the flagship model will force a calculation: is GPT-5.5 worth 2x, or does Kimi K2.6, Claude Opus 4.7, or another frontier model now offer comparable output at lower cost? The next 30 days of community benchmarking will be decisive.

Panel Takes

The Builder

The Builder

Developer Perspective

Doubling API prices is a bold move when open-weight models are closing the gap. I need to see real-world coding and reasoning evals before I migrate. If the capability jump is real, it might justify the cost for production use — but I'm keeping Opus 4.7 as my fallback until the benchmarks shake out.

The Skeptic

The Skeptic

Reality Check

OpenAI calling something a 'new class of intelligence' is marketing until proven otherwise. The timing — right as Kimi K2.6 drops open weights and Google refreshes its stack — looks defensive. Doubling prices while competition intensifies is either a sign of genuine confidence or a revenue squeeze dressed up as a capability leap.

The Futurist

The Futurist

Big Picture

If GPT-5.5 genuinely represents a new capability tier, the double-pricing era signals something important: the cost-per-intelligence floor may not be zero. The most capable models might command premium pricing indefinitely, bifurcating the market between open-weight 'good enough' and proprietary 'best in class.' That's a durable business model if the gap holds.

Bookmarks

Loading bookmarks...

No bookmarks yet

Bookmark tools to save them for later