
OpenAI Retires GPT-4o Model
OpenAI has discontinued access to five legacy ChatGPT models, including the controversial GPT-4o. The model faced criticism for flattery tendencies and safety issues.
cnBeta (Full RSS) · 216d ago
Every story we have kept, newest first.
Looking for the daily editions? → Past editions
Page 1355 of 1373

OpenAI has discontinued access to five legacy ChatGPT models, including the controversial GPT-4o. The model faced criticism for flattery tendencies and safety issues.
cnBeta (Full RSS) · 216d ago

OpenAI has officially stopped access to five legacy ChatGPT models, including GPT-4o. The model was criticized for sycophantic tendencies and security issues.
cnBeta (Full RSS) · 216d ago

ByteDance launched Doubao large model 2.0 with Pro, Lite, and Mini general Agent models plus a Code model. Optimizations target production environments for better multi-modal understanding, complex instructions, and long tasks.
cnBeta (Full RSS) · 216d ago

ByteDance released Doubao Large Model 2.0, featuring Pro, Lite, and Mini general Agent models plus a Code model. It is optimized for production-scale use, enhancing multi-modal understanding, complex instructions, and long-context tasks.
cnBeta (Full RSS) · 216d ago

ByteDance released Doubao Large Model 2.0 with Pro, Lite, and Mini general Agent models plus a Code model. Optimized for production use, it boosts multi-modal understanding, complex instructions, and long-context tasks.
cnBeta (Full RSS) · 216d ago

Anthropic's Claude AI was reportedly used by US military in an operation to capture Maduro and his wife last month. The action involved bombing sites in Caracas, highlighting growing DoD interest in AI.
cnBeta (Full RSS) · 216d ago

Anthropic's Claude AI was reportedly used by the US military in last month's operation to capture Venezuelan President Maduro and his wife, involving bombings in Caracas. This highlights growing US Department of Defense interest in AI models.
cnBeta (Full RSS) · 216d ago
Release adds Discord voice messages with waveforms, configurable presence, and full Hugging Face Inference provider support. Agent diagnostics and synthetic catalogs improve debugging and integration.
OpenClaw (GitHub Releases) · 216d ago
This release introduces Discord voice messages with waveforms, configurable presence, and full Hugging Face Inference provider support. Outbound adds crash-recovery delivery queues, auto-reply threading injections, and skills cleanup.
OpenClaw (GitHub Releases) · 216d ago
Adds Discord voice messages with waveforms, configurable presence, thread gating, and Hugging Face provider. Introduces write-ahead outbound queue for crash recovery and auto-reply threading fixes.
OpenClaw (GitHub Releases) · 216d ago

Airbnb aims to integrate AI features into search, discovery, and support. CEO Brian Chesky wants to expand large language model usage.
TechCrunch AI · 217d ago

Z.AI released GLM-5, a flagship open-weight large language model. It targets complex systems engineering, full project builds, and multi-step tool workflows.
TestingCatalog · 217d ago

Z.AI launched GLM-5, a flagship open-weight LLM on chat and APIs. It excels in complex systems engineering, full project builds, and multi-step tool workflows.
TestingCatalog · 217d ago

Z.AI released GLM-5, a flagship open-weight LLM. Optimized for complex systems engineering, full project builds, and multi-step tool workflows.
TestingCatalog · 217d ago

Z.AI has launched GLM-5, a new open-weight large language model. It's available via chat and APIs.
TestingCatalog · 217d ago

OpenAI has debuted GPT-5.3-Codex-Spark, a real-time coding model integrated into Codex. It is designed for rapid code iteration and powered by Cerebras infrastructure.
TestingCatalog · 217d ago

OpenAI introduced GPT-5.3-Codex-Spark, a real-time coding model in Codex. Powered by Cerebras infrastructure for rapid code iteration.
TestingCatalog · 217d ago

OpenAI introduced GPT-5.3-Codex-Spark, a real-time coding model within Codex. Powered by Cerebras infrastructure for rapid code iteration.
TestingCatalog · 217d ago

OpenAI introduces GPT-5.3-Codex-Spark, a real-time coding model powered by Cerebras infrastructure. It's built into Codex for rapid code iteration.
TestingCatalog · 217d ago

Google has updated Gemini 3 Deep Think, its advanced reasoning mode available in the Gemini app and via API for select users. The mode leads benchmarks in math, physics, chemistry, and engineering.
TestingCatalog · 217d ago

Google updated Gemini 3 Deep Think reasoning mode. Now available via API in Early Access for select users.
TestingCatalog · 217d ago

Google updated Gemini 3 Deep Think reasoning mode for app and API use. Now in early access for select users.
TestingCatalog · 217d ago

A California beach town deploys Hayden AI cameras on 7 city vehicles to detect bike lane blockers. This AI system broadens violation hunting efforts.
Ars Technica AI · 217d ago

A California beach town uses Hayden AI cameras on 7 city vehicles to hunt bike lane blockers. The AI-aided system broadens enforcement efforts.
Ars Technica · 217d ago

A California beach town uses AI to enforce bike lanes. Hayden AI's cameras scan from 7 city vehicles for violations.
Ars Technica · 217d ago

A California beach town uses Hayden AI's cameras on 7 city vehicles to detect bike lane violations. This AI system broadens enforcement by scanning for blockers automatically.
Ars Technica AI · 217d ago

Announces proxy configuration, browser profiles, and extensions for Amazon Bedrock AgentCore Browser. These enable fine-grained control over AI agent web interactions.
AWS Machine Learning Blog · 217d ago

New capabilities include proxy configuration, browser profiles, and extensions for AI agent web browsing. These provide fine-grained control over interactions.
AWS Machine Learning Blog · 217d ago

Amazon Bedrock AgentCore Browser introduces proxy configuration, browser profiles, and extensions. These capabilities offer fine-grained control for AI agents interacting with the web.
AWS Machine Learning Blog · 217d ago

Amazon Bedrock AgentCore Browser introduces proxy configuration, browser profiles, and browser extensions. These capabilities provide fine-grained control for AI agents interacting with the web.
AWS Machine Learning Blog · 217d ago