Nov 7, 2026  ·  1 min read

Google reroutes Gemini Embedding 2 Preview to the stable model

Google deprecated gemini-embedding-2-preview on November 6 and began routing its API requests to gemini-embedding-2 at the stable model's rates. Developers should update the model identifier. The replacement accepts text, images, video, audio, and PDFs, with an input limit of 8,192 tokens. Standard paid text input costs $0.20 per million tokens; other input types have separate rates.

The practical benefit is continuity for applications that use numerical representations of content for search or recommendations. Google says the replacement preserves the API schema, pricing, and embedding space. Existing stored representations therefore do not need to be rebuilt or re-indexed, and no code changes beyond the model string are required.

Source  ·  aistudio.google.com ↗ Google AI Studio The fastest path from prompt to production with Gemini
  1. Nov 8Vensum Power raises €9 million to scale software-controlled power converters
  2. Nov 8Uber details the trade-offs of guarding its production AI agents
  3. Nov 7Inkling 2 Small vs Tencent Hy4 vs DeepSeek V4.1 Flash 1017
  4. Nov 7Proposed test ties gravity's quantum nature to the mass of ticking clocks
  5. Nov 7WorkOS ties its data breach to a researcher's runaway AI agent

This is our sketchbook.

Two stoneware cups turned towards each other, as if in conversation

Got a good project?
Let’s work on it together.

Let’s talk →

info@ndot.io · New Jersey · Budapest · Biel