Global AI Digest

2,763 posts · Updated 31 min ago

Tracker Codex account reset Check whether Codex has reset today, and when it last did. Directory Free LLM APIs Every free AI model API tier, compared in one place. Briefing AI daily briefing The day's most important AI news, condensed into ten items.

Google Rolls Out Gemini 4 Carbon for Long-Horizon Agentic Workflows

Google is rolling out Gemini 4 Carbon, a frontier model built for deeply complex, long-horizon tasks across knowledge work, web development, and cyber defense,…

Google Gemini 31 min ago #Gemini 4 Carbon#Google#agentic workflows
Gemini 4 Carbon Lands in Cursor, Tops CursorBench at 60.4%

Cursor adds Gemini 4 Carbon, which reaches 60.4% on CursorBench at high effort, with 50% off pricing through November 30 before rates double on December 1.

Cursor 32 min ago #Gemini 4 Carbon#Cursor#CursorBench
Gemini 4 Carbon Now Available on Requesty at $1/$5 per Million Tokens

Requesty added Gemini 4 Carbon, a model aimed at multi-step coding, debugging and agent runs, priced at $1 per million input tokens and $5 per million output to…

Requesty 36 min ago #Gemini 4 Carbon#Requesty#pricing
Ripple 65-B Preview Now Available on Ollama for Agentic Coding

RicursiveAI's Ripple 65-B Preview is available on Ollama, built for agentic coding with up to 1M tokens of context and full zero data retention when run locally…

ollama 1 hour ago #Ollama#Ripple 65-B#agentic coding
Claude Fable 5.2 Launches with Early User Experiments

Anthropic released Claude Fable 5.2 and shared early experiments, including a code-built 3D cell model and a chalkboard-style step-by-step answer explainer.

Claude 1 hour ago #Claude Fable 5.2#Anthropic#AI model release
Muse Integrates Uber Courier for In-Chat Package Delivery

Muse, Meta's AI agent, can now arrange Uber Courier deliveries directly within the chat, letting users see pricing and approve bookings without leaving the conv…

Muse 2 hours ago #AI agent#Muse#Uber Courier
MiniMax Releases H3-Image With Open Weights, 16-Megapixel 4K Generation in 10-20 Seconds

MiniMax launched H3-Image, an open-weights image model that generates 16-megapixel (4K x 4K) images in 10-20 seconds, with strong instruction following, image e…

MiniMax (official) 2 hours ago #MiniMax#H3-Image#open weights
Command Code Adds Customizable Keyboard Shortcuts

Command Code now lets users remap keyboard shortcuts, switch models with a hotkey, and customize cursor navigation, viewable via the /hotkeys command.

Command Code 2 hours ago #Command Code#keyboard shortcuts#coding agent
OpenCode Uses Conservative Tokens-Per-Second Calculation, Causing Perceived Slowness

OpenCode computes tokens per second conservatively to prevent providers from inflating numbers, but since other harnesses use different methods, users report th…

dax 3 hours ago #OpenCode#tokens per second#LLM benchmarking
Expedia Becomes a Connector on Muse

Muse announced that Expedia is now available as a connector on its AI agent platform.

Muse 4 hours ago #Muse#Expedia#AI agent