AI news & updates
1,092 events
- Model launchMajorOpenAI releases GPT-6 Astra (low)
- Model launchMajorOpenAI releases GPT-6 Astra (medium)
- Model launchMajorOpenAI releases GPT-6 Astra (xhigh)
- Model launchMajorOpenAI releases GPT-6 Astra (pro, max)
- Model launchMajorOpenAI releases GPT-6 Astra (unknown thinking)
- AnnouncementSafety overview: GPT-6 Astra
GPT-6 Astra is our most capable broadly deployed model and our first to reach the Critical level of cybersecurity capability under our Preparedness Framework.
OpenAI✓ primaryOpenAI - AnnouncementFine-tuning a 350M Model for Better Structured Outputs in 100 GRPO StepsHugging Face✓ primaryHugging Face
- AnnouncementGive Your Coding Agents a Memory You OwnHugging Face✓ primaryHugging Face
- AnnouncementTraining a coding model to paint watercolours with TRL and OpenEnvHugging Face✓ primaryHugging Face
- PartnershipProactive cyber defense for governments and enterprises
The Fairwind Program is a limited access program for governments and trusted partners to use our cyber defense tools.
- AnnouncementATV Big Air Tour turned 3 days of work into 3 hours with ChatGPT
ATV Big Air Tour uses ChatGPT Work to speed up marketing, merchandising, and more. It even turned merchandise photos into an inventory website in 15 minutes.
- Model launchMajorMeta releases Muse Spark 1.3
Muse Spark 1.3 is a multimodal reasoning model from Meta for long-running agentic, multi-agent, and coding workflows. It is designed to keep track of information across extended tasks, work through...
- Model launchMajorGoogle releases Gemini 3.8 Flash
Gemini 3.8 Flash is Google's most intelligent Flash model with significant gains from 3.7 Flash across software engineering, agentic tasks, and multi-step reasoning.
- Model launchMajorGoogle releases Gemini 3.8 Flash (batch)
Gemini 3.8 Flash is Google's most intelligent Flash model with significant gains from 3.7 Flash across software engineering, agentic tasks, and multi-step reasoning.
- Model launchMajorMeta releases Muse Spark 1.3 Contributor
Muse Spark 1.3 Contributor is the cost-efficient contributor tier of Meta’s multimodal reasoning model for experimentation, learning, and early-stage agentic, multi-agent, and coding workflows. It is designed to track in
- Model launchMajorMeta AI releases Muse Spark 1.3 (xhigh)
- Model launchMajorGoogle DeepMind releases Gemini 3.8 Flash (unknown)
- Model launchMajorMeta AI releases Muse Spark 1.3 (unknown)
- Model launchMajorGoogle DeepMind releases Gemini 3.8 Flash (medium)
- Model launchMajorGoogle DeepMind releases Gemini 3.8 Flash (low)
- BenchmarkEpoch AI evaluates Qwen3.8 Max (0902) (xhigh)
GPQA diamond 92.3% · FrontierMath-Tiers-1-3-v2-Private 65.6% · FrontierMath-Tier-4-v2-Private 34.1% · OTIS Mock AIME 2024-2025 100.0%
- BenchmarkEpoch AI evaluates Gemini 3.8 Flash
GPQA diamond 95.4% · FrontierMath-Tiers-1-3-v2-Private 68.4% · FrontierMath-Tier-4-v2-Private 22.0% · OTIS Mock AIME 2024-2025 98.9%
- AnnouncementBenchMIRT: What are LLM benchmarks actually measuring?Hugging Face✓ primaryHugging Face
- FeatureThe latest AI news we announced in August 2026
Here are Google’s latest AI updates from August 2026
Google (The Keyword)✓ primaryGoogle - Model launchMajorIntroducing agentic video understanding with Gemini
- AnnouncementHow AI-native companies turn workflows into operating capability
Basis, Clay, and Exa Labs use AI agents to improve onboarding, account management, and developer integrations. See what enterprise leaders can apply.
OpenAI✓ primaryOpenAI - LaunchMajorTry Google Pics: Easy image creation and editing in Google Workspace
Built on our latest Nano Banana model, Google Pics — our image creation and editing tool — is now available.
- AnnouncementPath to Astra: critical capabilities and frontier safeguards
Astra is the first OpenAI model to meet the Critical cybersecurity capability threshold under the Preparedness Framework, with stronger safeguards for release.
OpenAI✓ primaryOpenAI - Model launchMajorQwen releases Qwen3.8 Max (0902)
Qwen3.8 Max 0902 is an updated snapshot of Qwen3.8 Max from Alibaba's Qwen team. It is a 2.4-trillion-parameter mixture-of-experts model that accepts text, image, and video input and returns text,...
- Model launchMajorAnthropic releases Claude Fable 5.1
Claude Fable 5.1 improves on Claude Fable 5 across the board, with the biggest gains in agentic coding, long-running agentic workflows, and knowledge work: long code refactors, front-end and visual...
- Model launchMajorAnthropic releases Claude Fable 5.1 (batch)
Claude Fable 5.1 improves on Claude Fable 5 across the board, with the biggest gains in agentic coding, long-running agentic workflows, and knowledge work: long code refactors, front-end and visual...
- Model launchMajorAlibaba releases Qwen3.8 Max (0902) (xhigh)
- Model launchMajorAnthropic releases Claude Fable 5.1 (xhigh)
- Model launchMajorAnthropic releases Claude Fable 5.1 (medium)
- Model launchMajorAnthropic releases Claude Fable 5.1 (low)
- Model launchMajorAnthropic releases Claude Fable 5.1 (unknown)
- BenchmarkEpoch AI evaluates Claude Fable 5.1
FrontierMath-Tiers-1-3-v2-Private 90.2% · FrontierMath-Tier-4-v2-Private 87.8% · OTIS Mock AIME 2024-2025 100.0% · SimpleQA Verified 70.8%
- LaunchMajorIntroducing @huggingface/kernels: 200+ WebGPU Kernels for Local AIHugging Face✓ primaryHugging Face
- Open releaseIBM releases Granite 4.2 8B
Granite 4.2 8B is a dense reasoning model from IBM. It is suited for mathematics, code generation, multilingual dialogue, and agentic workflows that need multi-step reasoning. It supports full, low-effort,...
- BenchmarkEpoch AI evaluates GPT-4o-mini
SimpleQA Verified 8.3%