Wirestack
Gemini tools can now return Office files, audio and video to the model (November 5, 2026)
Gemini function responses now take Office files, audio, video and Files API URIs. PPT, ODP and ODS files are accepted as input too.
Your keys
0 of 1 ready
Create each key on its own page, paste it below, then copy the finished lines into your project's .env file. What you paste stays in this tab. It is never sent to Wirestack or saved.
- Google Gemini API API keyGEMINI_API_KEYGet key
Click Create API key
3 of 3 free copies left this month.Unlimited with Pro
What you get
On November 5, 2026 Google changed what tools can send back to Gemini. A function response can now carry Office and OpenDocument files, audio and video, as well as images, PDFs and plain text. A tool can also return a Files API URI instead of inline bytes.
- Interactions API.
function_resultsteps now acceptdocument,audioandvideocontent blocks next totextandimage. Each block takes base64dataor auri. Video blocks work inv1betaandv1alphaand acceptprocessing("static"or"agentic"). - GenerateContent API. Function response parts accept audio, video and Office and OpenDocument files. They also accept
fileData.displayNameworks on bothinlineDataandfileData, and video parts acceptmediaProcessing(STATICorAGENTIC). - More input formats. The Gemini API now takes PPT, ODP and ODS files, and renders DOC files as page images. DOC, PPT and ODP pages are read as images. For ODS, the model gets the data and formulas from the first sheet only.
Files a tool returns are processed the same way as the same file sent as normal input. An unsupported MIME type returns 400 INVALID_ARGUMENT.
Prerequisites
- A Gemini API key in
GEMINI_API_KEY - The Google Gen AI SDK for Python (
google-genai) 2.33.0 or newer. That release added audio, document and video content to function results. - A Gemini 3 series model. The docs examples use
gemini-3.8-flash - An earlier interaction in which the model called your tool (the object below)