Dicas
5 min
Tencent's Hy4: 1M context, vision and tool calling
Tencent's Hy4 brings 1 million tokens of context, vision and native tool calling for code agents and long duration tasks.
#tencent#hy4#agentes
Artificial Intelligence
Blog
Articles, analyses and tutorials on artificial intelligence, autonomous agents and software development.
Newsletter
Get AI articles in your inbox
Technical curation on models, agents and coding tools. No spam, unsubscribe anytime.
Dicas
See all: Dicas
Dicas
6 min
5 tips to operate desktop with agent-desktop
Vercel Labs launched agent-desktop, a tool that lets AI agents operate desktop applications via the accessibility tree. See how it works.
#radar#ia
Dicas
5 min
How to track Codex CLI releases without breaking your workflow
The openai/codex repository publishes releases frequently. See how to keep up with new versions without putting your workflow at risk.
#Codex CLI#GitHub#releases
Dicas
6 min
5 tips for using Codex CLI in your workflow
The official openai/codex repository hosts Codex CLI. Here are 5 practical tips to use the tool with more control in your development workflow.
#Codex#CLI#OpenAI
Dicas
5 min
How to use SuperCLUE to choose an AI model
The November 2026 SuperCLUE covers 19 models and category leaders. This guide shows how to use the category view to pick the right model.
#radar#ia
Dicas
6 min
Claude Code 2.1.319: Background MCP and Permissions
Version 2.1.319 of Claude Code adds background MCP tasks, the permissions.denyInsteadOfPrompting option, and new diagnostic commands.
#radar#ia
Benchmarks
See all: Benchmarks
Benchmarks
6 min
SuperCLUE November 2026: different leaders by category
SuperCLUE published in November 2026 a table with 19 models and different leaders in each category. The split ranking is more useful as a shortlist map than as a final verdict.
#benchmark#SuperCLUE#modelos
Benchmarks
7 min
PLE: physical engineering benchmark for AI agents
Google Cloud launched the Physical-Engineering Last Exam, an 85 task physical engineering benchmark. Every model scored below 50%. See the numbers.
#radar#ia
Benchmarks
5 min
BenchLeader November 7: Claude Opus 5.5 tops the aggregated benchmark index
BenchLeader updated its index on November 7, 2026. Claude Opus 5.5 leads with 69.3 points, closely followed by Gemini 4 Carbon and GPT-6 Astra.
#benchmark#BenchLeader#Claude Opus 5.5
Benchmarks
5 min
Ripple 65B Preview tops TerminalBench 2.1 among U.S. open models
Ricursive says the Ripple 65B Preview achieves the highest score among U.S. open models on TerminalBench 2.1, with a cost of $32 million.
#benchmark#modelos abertos#Ripple 65B
Benchmarks
4 min
VIALS-AA: Benchmark for Reasoning on Scientific Images
Artificial Analysis launched VIALS-AA, an implementation of Handshake AI's VIALS benchmark, which tests whether models can interpret scientific images and answer based on the evidence.
#radar#ia
Benchmarks
5 min
Inkling 2 Small: benchmark shows strength in enterprise tasks
Thinking Machines Lab announced Inkling 2 Small with 594 billion parameters and 17 billion active. The model scored 71.8% on EnterpriseBench-CoreCraft and 46.8% on EnterpriseOps-Gym.
#radar#ia
Newsletter
See all: Newsletter
Newsletter
6 min
GitHub rebuilds Git storage for agents
GitHub announced its new Git storage architecture, Project Rubicon, capable of handling thousands of writes per second to a single repository.
#github#git#infraestrutura
Newsletter
7 min
Palette: IBM model learns to create slides in your style
IBM and Hugging Face released Palette, a 20 billion parameter model that generates presentations in code trained to follow the IBM Design Language.
#IA#modelos#apresentações
Newsletter
5 min
OpenAI creates a standard for responsible AI infrastructure
OpenAI published a standard for responsible AI infrastructure with commitments on energy, water, workforce and communities.
#radar#ia
Newsletter
6 min
Cult.fit standardizes operations with Codex and builds internal apps
Cult.fit standardized development on Codex, built more than 20 internal apps with non-technical teams and cut support costs by over 50%.
#radar#ia
Newsletter
5 min
Meta launches Spaces for the Muse agent
Meta launched Spaces for the Muse agent, creating persistent workspaces that update automatically. See how it works.
#radar#ia
Newsletter
6 min
Perplexity launches Teammates, an always-on agent for work
Perplexity launched Teammates, a persistent AI agent that takes on continuous tasks. See how it works, the cost, and how it compares to Dots and Duet.
#radar#ia