Google released the Gemini Decisions API to score up to 1,000 questions in one pass. No word on how teams audit the confidence scores returned by the model.
Startup Radical Numerics has developed an AI model designed to identify both natural and AI-engineered pathogens without requiring an internet connection.
Today we’re launching the Gemini Decisions API, powered by our new Gemini Decision 1 model in @GoogleAIStudio.
Instead of generating text token by token, it’s built for closed-set choices scoring up to 1,000 questions across text, images, audio, video + PDFs in a single pass. A few details:
- 3 decision types: yes/no, multiple choice (up to 255 options) + rubric scoring
- $0.10 / 1M input tokens with $0 billed for output tokens (~5x cheaper than Flash-Lite)
- leads on several benchmark evals with lowest calibration error
Google teams are already building with it, below is a look at Stitch Rapid designing a full product flow in seconds (1x speed)
Introducing Gemini Decision 1, our new frontier decision model in the Gemini API.
It is $0.10 per 1M input tokens with no additional cost. With support for image, audio, video, and PDF natively, 256K context window, & 1K questions.
You can try it for free in the API!
A ChinaTalk investigation reported by The Decoder states US data brokers sold roughly $500 million in reinforcement learning training environments to Chinese AI labs, including datasets built for OpenAI and Anthropic.
Security firm Claroc found 35,578 out of 37,571 Grok-built apps were open to the public without a login. While ChatGPT defaults newly created shared apps to Only me, SpaceXSI preselected Anyone with the link on paid Grok accounts.
Google DeepMind announced AlphaGenome 2, a DNA sequence model that conditions on cell types and caught 30% of causal variants on the TenK10K dataset. They haven't shared what commercial pricing will look like once it leaves private preview.