Gemma 4 26B A4B IT: an official availability record, not a filled-in spec sheet
Google published the exact model ID and where to find it. Context, pricing, and benchmark figures are still missing.
By the benchr team · · Changelog · Provider-published facts rechecked against the official sources on July 28, 2026
GOOGLE · GEMMA426B A4B · OPEN
Benchr model field plateGemma 4 26BOpen record · verified gaps
Editorial imageA BenchR editorial illustration of an open model record: verified facts are connected, and unreported fields remain visibly unfilled.
Released2 Apr2026
Model ID26BA4B IT
Context—Not published
API price—Not published
Google released gemma-4-26b-a4b-it on April 2 as part of the Gemma 4 launch. The record is short because Google's public details for this exact ID are short.
The model ID is the useful part
gemma-4-26b-a4b-it is the string you can look for in AI Studio and the Gemini API. Confirm that your project can see it before you design around it. A copied name from a launch note is not the same as account-level access.
Blank fields are still information
The release note does not list a context window, per-token API price, or benchmark result for this exact model. benchr will not copy those fields from a similarly named Gemma, an older card, or a third-party listing. Treat each blank as a question for Google, not a favorable assumption.
This is an availability note, not a review
The number in the name is not a deployment specification. Until Google publishes a model-specific card or price table, the useful record is what exists, where it appears, and what a production team still needs to learn.