Every change to the prices in our dataset, with the date it was made. This list is generated from the version history of the data file itself, so it cannot leave anything out. The current data was last checked on against each provider’s official pricing page; the next full recheck is due in mid-November 2026, and again right after 1 January 2027, when Gemini 3.8 Flash’s announced price increase takes effect.
Price changes
19 models checked, 19 of them on the provider’s official page. No price changed on models already listed.
Google Gemini 4 Argon added: $2 input, $10 output, $0.1 cached input per 1M tokens.
18 models checked, 18 of them on the provider’s official page. 6 price fields changed.
Anthropic Haiku 4.5 removed from the table (was $1 input, $5 output).
DeepSeek (current flagship) removed from the table (was $0.28 input, $0.42 output).
First version of the dataset: 13 models, 7 confirmed on the provider’s official page and 6 cross-checked across secondary sources because the official page blocked automated access at the time.
Gemini 4 Argon: launch price until 30 November 2026; from 1 December 2026, $4 input and $20 output per 1M tokens (Google’s pricing page).
Gemini 3.8 Flash: from 1 January 2027, $1.50 input and $7.50 output per 1M tokens (Google’s pricing page).
OpenAI GPT-5.6 Sol: promotional price, stated as available at least through 21 November 2026 (OpenAI’s pricing page).
Site updates
— Addresses that do not exist now show a proper “page not found” page in your language, with links to the calculator and price pages, instead of a copy of the home page.
— Prices rechecked on all five official pages. New: four long-form guides (tokens, caching and batch discounts, monthly cost estimates, input vs output pricing), this price history, cache-write and batch-discount data per model, a fuller About page with our checking method, and method, common mistakes and sources on the calculator and comparison pages.
— Mobile fixes after a 360px check: tables scroll sideways, filter buttons wrap, larger tap targets, side padding restored.
— Page titles reworded to match what people search for.
— Provider pricing pages for Claude, OpenAI, Gemini, Grok and DeepSeek in all seven languages, each with monthly costs for typical workloads; fixed an invisible mobile menu icon.
— Price recheck: every row now confirmed on the provider’s own page. Share image and app icons added; titles and descriptions shortened.
— Privacy policy in seven languages, a contact address on every page, and Google’s consent message for visitors in the EEA, UK and Switzerland.
— Every page pre-rendered as plain HTML; worked example, FAQ and embed code added under the calculator and comparison table.
— Unknown addresses now return a real 404; redirect chains removed.
— The site became available in seven languages: English, Spanish, Italian, French, German, Chinese and Greek.
— QuotaCost launched: monthly cost calculator and sortable price comparison, every price sourced and dated.
Found a price that is wrong or out of date? Email contact@quotacost.com with the model and the provider page you are looking at. Corrections show up here with their date.