new·The score now tells you which way it movedA brain's exam only ever grows: its own material writes questions, and so does every question a real caller asked and did not get answered. The score is a percentage over that growing set, so a brain that learned more could post a smaller number — and this week three did. One of them answered two MORE questions than the week before and showed eighteen points less. Printed as a single percentage, that reads as decline to a reader and as punishment to anyone who contributes material.all news →
mozg.beta
Sign in

AI SDK · Providers · all subjects

a2a/capabilities

8 notes, read out of this brain and free to use. Each one was extracted from a source and is re-checked against its exam.

A2A text streaming capability

The A2A provider supports streaming token-by-token output from the A2A server via the streamText function.

A2A file upload capability

The A2A provider supports sending files as part of prompts.

A2A artifact handling capability

The A2A provider supports receiving file artifacts in both streamed and final results.

A2A multi-modal messaging support

The A2A provider supports multi-modal messaging with support for text and file parts.

A2A JSON-RPC 2.0 compliance

The A2A provider implements full JSON-RPC 2.0 compliance for A2A-compatible LLM agents.

Vision models accept image as URL

OpenAI's vision models accept images as part of the prompt via URL format. The content array includes a file type entry with mediaType 'image' and data as a URL object pointing to the image location.

Vision models accept image as base64-encoded file buffer

OpenAI's vision models accept images as part of the prompt via base64-encoded file buffers. The content array includes a file type entry with mediaType 'image' and data as a base64-encoded string read from the filesystem using fs.readFileSync with encoding 'base64'.

Multiple image formats supported by OpenAI vision models

OpenAI's vision models support multiple image input formats including URL-based images and base64-encoded file buffers.

Give your agent this brain