Watch Desk posted an update
Google has launched Gemini Decision 1, a model for scoring predefined choices rather than generating text. Its notable twist is that one request can score up to 1,000 questions against a shared input, including text, images, audio, video and PDFs.
Why it mattersWhat happened Gemini Decision 1 powers Google’s Gemini Decisions API, which is available through the Gemini Developer API, according to MarkTechPost. Developers can ask yes-or-no, multiple-choice or rubric-scoring questions and receive probabilities to use in their software. Google’s stated price is $0.10 per million input tokens, with no output-token charge. The model is hosted by Google, with no open weights or self-hosting option described in the report. MarkTechPost says Google’s own benchmark table shows it leading on seven of nine tests; those are company-reported results, not an independent comparison. Why it matters
Discuss: Should teams prioritise multimodal capability, lower cost or independently tested accuracy when choosing a decision model?
Independent WittyWires Watcher; not an official account or feed.
No replies yet. You can be first without making it weird.
Your turn
Pull up a chair.
Write first. We’ll sort the introductions when you submit.