OpenAI's price war gambit: GPT-5.6 Sol tokens get 20% cheaper for 90 days

Share
OpenAI's price war gambit: GPT-5.6 Sol tokens get 20% cheaper for 90 days

OpenAI picked the eve of its rival's roadshow to cut developers a deal. A three-month, 20-percent-plus discount on its flagship model — announced hours after Anthropic's bankers floated a $2 trillion IPO — is business warfare conducted in a price list.

OpenAI said Friday it is cutting API and credit prices for GPT-5.6 Sol by more than 20 percent for the next three months, taking the frontier model to $4 per million input tokens and $20 per million output tokens. The company framed it as efficiency gains — pricing falls "as we make it more efficient to run" — with subscription usage untouched and credits stretching further across ChatGPT Work and Codex token-based plans. But the timing does the talking: it landed hours after reports that Anthropic's bankers have told potential investors its IPO could raise $100 billion-plus at a $2 trillion valuation, and analysts immediately framed the cut as pressure on Anthropic's enterprise pitch while that roadshow narrative is at its most valuation-sensitive.

The move also reads as a shot across an industry that has been drifting toward price discipline. Chinese labs are already testing preemptive reductions on new models like GLM-5.3-Flash, and a declared 90-day window from the priciest name in the market sets an expiry date everyone else now plans around.

The economics explain why OpenAI can afford it. Frontier inference carries fat incremental margins — one widely shared estimate puts serving Sol on a B200 cluster around 80 percent gross margin — so a temporary cut costs little and buys competitive position. It compounds a revenue story we covered earlier this month — GPT-5.6 Sol powers OpenAI's revenue rebound over Anthropic — where Ramp's spending index already showed OpenAI outgrowing Anthropic 82-to-76 percent in Q3 after briefly losing the lead. Cheaper tokens make that gap harder for Anthropic to close back.

What to watch: whether Anthropic answers with price moves during the roadshow or holds pricing to protect the margin story its filing needs — and what happens when the 90-day truce lapses in November.

Would you switch model vendors over a 20% discount, or does switching cost beat any price cut?

Read more

Mistral unveils Le Chonk: a 1T-parameter open-weights model

Mistral unveils Le Chonk: a 1T-parameter open-weights model

The biggest open-weight release outside China lands in public preview today, and the country that spent the week promising its own frontier model just put a price on the ambition. Mistral has opened a public preview of Mistral Large 4 — codenamed "le Chonk" — a 1 trillion-parameter mixture-of-experts model with 49 billion active parameters, natively multimodal, which the company calls its largest and most capable model to date. The preview API is live today on Mistral Studio at $1.36 per milli

Google signs 3.6 GW power deal, a quarter of it new nuclear

Google signs 3.6 GW power deal, a quarter of it new nuclear

Grid capacity, not chips, is becoming the binding constraint on the AI buildout — and on the same day, the labs told an Australian inquiry they can live with mandatory incident reporting. Google has contracted 3.6 GW of power from Constellation Energy across the PJM grid — the largest electricity deal in the region's history, and the biggest single power commitment any AI company has made. The agreement covers 3,590 megawatts over 13 states, with 890 megawatts of new nuclear capacity coming fr

Korea probes AI agents in bank hacks as president cites 'signs'

Korea probes AI agents in bank hacks as president cites 'signs'

South Korea opened a formal investigation into whether AI agents drove a wave of bank breaches — and it isn't the only AI story moving money today. President Lee Jae Myung said "signs" point at AI models, and the probe is now at the highest level a national banking sector has seen. Speaking at a cabinet meeting, Lee said that "in some hacking incidents, signs have emerged of AI being used, causing considerable public concern and anxiety," and police have since opened a full-scale investigation

Open Source Radar — October 6: nothing leaves your machine

Open Source Radar — October 6: nothing leaves your machine

Today's trending board is all projects we ran earlier this week, so the fresh signal comes from the Product Hunt launch slate instead — three open-source tools that share one instinct: your phone, your pixels and your MCP traffic should stay on hardware you control. All three verified at the source. iphone-use (Rust, MIT, about 59 stars) is computer-use for a real iPhone: an agent reads the screen as text, taps, swipes and types over WebDriverAgent, and every action comes back with an honest v