Skip to content
Hugin
A person uses a mouse behind an open laptop at a wooden table.

Hugin News

Google lists December rates for Gemini 4 Carbon while OpenAI dates its Sol promotion

2 min read

AI-generated editorial illustration for Hugin.

This record was prepared on the date shown and recovered after a publishing interruption.

Google's Gemini 4 Carbon pricing table gives December 1 input and output rates for four processing modes. OpenAI lists November 30 for the end of GPT-5.6 Sol's promotion without giving a December replacement rate on its pricing or model page. Three other OpenAI models have a separate promotional commitment that runs at least through December 31.

googleopenaigeminiapi-pricingdevelopers
source receipts
5
source hosts
2
read time
2min
primary source
Linked

A developer budgeting for December can look up the new per-token rates for Google's Gemini 4 Carbon today. OpenAI gives an end date for its GPT-5.6 Sol promotion, but its pricing and Sol model pages give no replacement rate for December.

Carbon's rates depend on processing mode

Google's November 10 launch announcement gave introductory prices of $2 per million input tokens and $10 per million output tokens through November 30. Its pricing table identifies those as Standard rates. Starting December 1, Standard input costs $4 per million tokens and output costs $20 per million, including tokens the model uses for thinking.

Batch and Flex each move from $1 for input and $5 for output to $2 and $10 respectively, per million tokens. Priority moves from $3.60 for input and $18 for output to $7.20 and $36. Google also lists distinct prices for implicitly cached input. Under Standard processing that rate changes from $0.10 to $0.20 per million tokens. Carbon has no free tier.

Sol's announcement names a date

OpenAI's August 21 changelog set a minimum duration for GPT-5.6 Sol's promotion, at least through November 21. The pricing page and Sol model page now give November 30 as the end of the promotional window. The model page lists $4 per million input tokens, $0.40 per million cached input tokens and $20 per million output tokens. For prompts with more than 272,000 input tokens, it lists twice the input rate and 1.5 times the output rate for the full request. The pages do not specify how much a comparable Sol request will cost once that promotion ends.

OpenAI separately says promotional pricing for GPT-6 Astra for Law, GPT-6.2 Astra for Law and GPT-6 Rosalind will remain available at least through December 31. That wording leaves room for the prices to continue into January.

Disclosure: OpenAI's Codex assisted with reporting this article.

Source links