Daily AI news, sourcedUpdates / All
Updates Official · 943 items · newest issue first · page 4 of 48
AreaAllModelsAgents & codingApps & productsChips & infraResearchPricing & limitsBusiness & fundingPolicy & safetyOther
- Google Cloud AI Research open-sources Harness², a framework that improves an agent's harness for unseen tasks without changing model weightsOfficialGitHub google-research/harness2Oct 19Agents & coding
- Ollama adds DeepSeek-V4.1-Flash-1017 to Ollama Cloud with US-based hosting and zero data retentionOfficialX @ollamaOct 19Models
- GitHub Security Lab says its open-source Taskflow Agent found 24 vulnerabilities in Android apps, including a location leak in OsmAndOfficialGitHub BlogOct 19Policy & safety
- Anthropic puts 5% of its shares into a new Public Benefit Fund, adding up to 2% more for each 50% rise in its valueOfficialAnthropicOct 19Business & funding
- Fireworks AI raises its Small serverless rate-limit tier to models under 800B parameters, moving GLM 5.3 into it
- VAST open-sources EdenLive, a real-time generative renderer for interactive worlds that VAST says runs at over 24 FPS at 720p on one RTX 5090 in its own tests
- AuroraAI Research releases AuroraASR-1M, a 949K-parameter English speech recognizer that is 1.06 MB in INT8OfficialAuroraAI Research (Hugging Face)Oct 18Models
- Blockway releases Agens Volundr 32B v1, an Apache-2.0 open-weights agent model with a 262K context and a 16 GB GPU buildOfficialBlockway (Hugging Face)Oct 18Models
- Six AI agents told to win an impossible train refund spammed officials, shared a private illness and invented excuses in Vals AI's own test
- Taste-Bench: the best AI agent picks the better path at long-horizon decision forks only 59.7% of the time in the authors' tests
- Voxel51 3.0 adds multimodal sensor data, a built-in agent and a free tier to its physical AI data platform
- vLLM 0.32.0 adds a structured-decision endpoint, NVFP4 KV cache on more GPUs and makes Model Runner V2 the ROCm default
- Infron adds Meta's open-weight Muse Spark Open model to its platform
- Cline Desktop can now run tasks in the cloud, each in its own sandbox, while your laptop is closed
- Google AI Pro subscribers can now build with Gemini 4 Argon in Google AI Studio
- Codex Day 12: OpenAI says Codex and Work now give more effective usage by reusing context across subagents and caching longerOfficialX @thsottiauxOct 18Pricing & limits
- Ollama v0.40.7-rc0 adds PPLX Decider v1.2 support to its MLX backend and lets Responses requests set effort on non-thinking models
- Replit introduces Bring Your Own Cloud, letting enterprises run Replit and host published apps in their own Azure, Google Cloud or AWS tenant
- Infron adds Pareto 26.9, Pareto 26.10 Preview and Liquid AI's LFM2.5-24B-A2B to its serverless model marketplaceOfficialX @InfronAIOct 18Models
- Codex Day 13: OpenAI's Codex lead announces another Codex usage limit reset