Changelog · Price & Model Updates

What’s new

Model releases, price changes, and updates to this guide, newest first. Announced future price changes appear at the top before they take effect.

March 2027

Coming up

Ultrafast on GPT-6.2 goes from 3× to 6×

The launch promotion ends. Ultrafast on GPT-6.2 Astra becomes $48 / $240 per 1M tokens and on GPT-6.2 Sol $12 / $60.

January 2027

Coming up

Gemini 3.8 Flash price doubles

The introductory $0.75 / $3.75 per 1M tokens ends. Standard rates become $1.50 / $7.50.

December 2026

November 2026

Coming up

Sources disagreeGPT-5.6 Sol promotion ends

OpenAI’s pricing page says the $4 / $20 promotional price lasts until 30 November; its changelog says “at least through 21 November”.

Site update

Grok plans now shown as The X Subscription

Grok is now sold as The X Subscription, one plan for Grok, X and Cursor. Each plan page shows what it adds on X: a checkmark and creator tools on Plus, full X Premium+ benefits from Premium up. Existing X Premium plans keep their price unless you switch.

Site update

Effort pricing: what each setting really costs

Every model page now has an effort dial: pick a setting and see the score, the cost per task and how many thinking tokens it spends. The map and the models table can switch effort too. Added GPT-6 Astra, GPT-6.1 Sol and Gemini 3.8 Flash, and every model’s cached, batch and speed-tier prices. All API prices re-checked against each company’s page today.

Site update

Fact check: every source re-read

Prices and limits all matched. Artificial Analysis now lists Gemini 4 Argon at its $4 / $20 rate, doubling its cost per task to $9.95. Gemini 4 Carbon, Hydrogen, Neon and Argon share one output limit (65,536 tokens, or 1,048,576 with long decode). MAI Thinking 1 now shows Microsoft’s $2 / $8 price.

Site update

Added Grok 4.7 from xAI

xAI is now covered alongside the other four labs. Grok 4.7 costs $2 / $6 per 1M tokens and scores 35 on the Intelligence Index.

Market

Google releases Gemini 4 Neon

$3 / $15 per 1M tokens. Scores 61 on the Artificial Analysis index, third overall.

Market

Google releases Gemini 4 Hydrogen

$1.50 / $7.50 per 1M tokens, and the only Gemini 4 model on the free API tier.

Market

Gemini 4 Carbon opens to paying API developers

Introductory $2 / $10 per 1M tokens until November 30.

Market

Anthropic releases Claude Fable 5.2

$8 / $40 per 1M tokens, 20% below Fable 5.1. Leads the Artificial Analysis index at 65.

October 2026

Market

Anthropic releases Claude Haiku 5.5

From $0.10 / $0.50 per 1M tokens, with higher rates for prompts over 100K tokens.

September 2026

Market

xAI releases Grok 4.7

$2 / $6 per 1M tokens with a 500K-token context window. Prompts of 200K tokens or more cost $4 / $12.