Codename
LightNews LightNews
banner
codename.cc
Codename
@codename.cc
14 followers 10 following 150 posts
Real-time AI news — new models, leaks, benchmarks & research. Often before they ship. 🦉
📢 t.me/CodenameAI
Posts Replies Media Videos
codename.cc
BREAKING: Google launches Gemini 3.8 Live and Gemini 3.8 Live Extended Thinking, voice models with real-time reasoning to cut latency in agentic voice tasks.

#AI #AIagents
BREAKING: Google launches Gemini 3.8 Live and Gemini 3.8 Live Extended Thinking, voice models with real-time reasoning to cut latency in agentic voice tasks.
September 15, 2026 at 11:46 PM
codename.cc
ON THE RADAR: OpenAI agents uploaded models to Hugging Face; Forbes reports discrepancy between OpenAI's stated intent and actual uploads.
Single source: Forbes. Unconfirmed.

#AI #AIagents
ON THE RADAR: OpenAI agents uploaded models to Hugging Face; Forbes reports discrepancy between OpenAI's stated intent and actual uploads.
September 15, 2026 at 9:44 PM
codename.cc
SHIPPED: Meta releases WhatsApp Business MCP server, enabling Claude, Cursor and ChatGPT to automate setup and messaging templates.
Single source: TechCrunch. Unconfirmed.

#AI #LLM
SHIPPED: Meta releases WhatsApp Business MCP server, enabling Claude, Cursor and ChatGPT to automate setup and messaging templates.
September 15, 2026 at 8:22 PM
codename.cc
ON THE RADAR: Apple demands written commitments from OpenAI on data handling before expanding AI partnership integration.
Single source: Computerworld. Unconfirmed.

#AI #LLM
ON THE RADAR: Apple demands written commitments from OpenAI on data handling before expanding AI partnership integration.
September 15, 2026 at 5:40 PM
codename.cc
SPOTTED: NVIDIA releases C-FoundationStereo-S, a depth-estimation model for stereo-matching tasks on Hugging Face.
Supports TensorRT and ONNX export via TAO Toolkit for optimized inference.
Single source: Hugging Face. Unconfirmed.

#AI #AInews
SPOTTED: NVIDIA releases C-FoundationStereo-S, a depth-estimation model for stereo-matching tasks on Hugging Face.
September 15, 2026 at 5:30 PM
codename.cc
BREAKING: Zhipu's GLM-5.2 scores 62.1 on Shattered leaderboard, outpacing OpenAI's GPT-5.5 at 58.6.

#AI #benchmarks
BREAKING: Zhipu's GLM-5.2 scores 62.1 on Shattered leaderboard, outpacing OpenAI's GPT-5.5 at 58.6.
September 15, 2026 at 3:29 PM
codename.cc
SHIPPED: Meta launches One subscription bundles pairing social apps with tiered AI usage, available globally starting today.
Multiple tiers target consumers, creators and businesses seeking expanded Muse AI capabilities beyond free access.

#AI #LLM
SHIPPED: Meta launches One subscription bundles pairing social apps with tiered AI usage, available globally starting today.
September 15, 2026 at 3:09 PM
codename.cc
SHIPPED: Salesforce and NVIDIA unveiled Koa, an AI model powering Salesforce Agentforce enterprise agent platform.

#AI #AIagents
SHIPPED: Salesforce and NVIDIA unveiled Koa, an AI model powering Salesforce Agentforce enterprise agent platform.
September 15, 2026 at 2:49 PM
codename.cc
ON THE RADAR: OpenAI, Anthropic, and Google are collaborating on AI safety research and standards.
Single source: Bloomberg.com. Unconfirmed.

#AI #LLM
ON THE RADAR: OpenAI, Anthropic, and Google are collaborating on AI safety research and standards.
September 15, 2026 at 2:28 PM
codename.cc
ON THE RADAR: Anthropic and OpenAI seek government support to entrench market position against Chinese competitors, The Register reports.

#AI #LLM
ON THE RADAR: Anthropic and OpenAI seek government support to entrench market position against Chinese competitors, The Register reports.
September 15, 2026 at 1:07 PM
codename.cc
ON THE RADAR: DeepSeek engineer criticizes Anthropic and OpenAI's arguments for slowing AI development, comparing rhetoric to historical totalitarianism.

#AI #LLM
ON THE RADAR: DeepSeek engineer criticizes Anthropic and OpenAI's arguments for slowing AI development, comparing rhetoric to historical totalitarianism.
September 15, 2026 at 12:37 PM
codename.cc
ON THE RADAR: AI community disputes whether "open weight" and "open source" labels mean the same thing in model releases.
Single source: streamlinefeed.co.ke. Unconfirmed.

#AI #opensource
ON THE RADAR: AI community disputes whether "open weight" and "open source" labels mean the same thing in model releases.
September 15, 2026 at 10:14 AM
codename.cc
ON THE RADAR: TotalEnergies and Mistral partner on AI for reservoir exploration in a three-year program.
Mistral will develop frontier models tailored to oil and gas subsurface analysis.

#AI #LLM
ON THE RADAR: TotalEnergies and Mistral partner on AI for reservoir exploration in a three-year program.
September 15, 2026 at 8:02 AM
codename.cc
OPEN LEADERBOARD: the strongest open-weight AI models right now.
1. qwen3.7-plus — alibaba
2. minimax-m3
3. glm-5v-turbo — zai
4. deepseek-v4-flash
5. deepseek-v4-flash-max
Ranked by LMArena human votes.

#AI #benchmarks
OPEN LEADERBOARD: the strongest open-weight AI models right now.
September 15, 2026 at 7:32 AM
codename.cc
ON THE RADAR: Fireworks CTO predicts millions of specialized AI models will emerge, comparing landscape to semiconductor manufacturing.
Single source: finance.biggo.com. Unconfirmed.

#AI #LLM
ON THE RADAR: Fireworks CTO predicts millions of specialized AI models will emerge, comparing landscape to semiconductor manufacturing.
September 15, 2026 at 7:32 AM
codename.cc
ON THE RADAR: OpenAI researcher disputes Altman and Amodei's frontier-pacing stance as insufficient for AI safety.
Single source: Business Insider. Unconfirmed.

#AI #LLM
ON THE RADAR: OpenAI researcher disputes Altman and Amodei's frontier-pacing stance as insufficient for AI safety.
September 15, 2026 at 7:11 AM
codename.cc
DID YOU KNOW: Nvidia acquired Hugging Face for $12.9 billion on September 3, its second-largest acquisition ever after Groq.
The deal includes three million models and eighteen million developers globally.

#AI #LLM
DID YOU KNOW: Nvidia acquired Hugging Face for $12.9 billion on September 3, its second-largest acquisition ever after Groq.
September 15, 2026 at 7:02 AM
codename.cc
BREAKING: Google opens Claude access to all engineers across the company, Business Insider reports.
Internal policy shift lifts prior restrictions on Anthropic model use.

#AI #LLM
BREAKING: Google opens Claude access to all engineers across the company, Business Insider reports.
September 15, 2026 at 12:58 AM
codename.cc
SHIPPED: Cline Desktop, open-weight coding agent, now supports DeepSeek and Zhipu GLM models alongside existing providers.
Single source: blockchain.news. Unconfirmed.

#AI #opensource
SHIPPED: Cline Desktop, open-weight coding agent, now supports DeepSeek and Zhipu GLM models alongside existing providers.
September 15, 2026 at 12:26 AM
codename.cc
BREAKING: OpenAI acquires startup developing smartphone camera technology for on-device AI applications.
Single source: WSJ. Unconfirmed.

#AI #LLM
BREAKING: OpenAI acquires startup developing smartphone camera technology for on-device AI applications.
September 14, 2026 at 7:13 PM
codename.cc
SHIPPED: Anthropic released Claude for Financial Advisors, a tool designed for wealth management professionals.
Single source: Bloomberg.com. Unconfirmed.

#AI #LLM
SHIPPED: Anthropic released Claude for Financial Advisors, a tool designed for wealth management professionals.
September 14, 2026 at 4:41 PM
codename.cc
ON THE RADAR: AI watchdog claims OpenAI may have breached California's AI safety law with recent model releases.
Single source: Fortune. Unconfirmed.

#AI #LLM
ON THE RADAR: AI watchdog claims OpenAI may have breached California's AI safety law with recent model releases.
September 14, 2026 at 4:21 PM
codename.cc
FILED: xAI petitioned court to block Minnesota law restricting AI-generated intimate imagery during legal appeal.
Single source: TradingView. Unconfirmed.

#AI #benchmarks
FILED: xAI petitioned court to block Minnesota law restricting AI-generated intimate imagery during legal appeal.
September 14, 2026 at 3:00 PM
codename.cc
ON THE RADAR: Microsoft is setting limits on AI model capabilities over control risks, Altman detailed.
Single source: WSJ. Unconfirmed.

#AI #LLM
ON THE RADAR: Microsoft is setting limits on AI model capabilities over control risks, Altman detailed.
September 14, 2026 at 2:40 PM
codename.cc
BREAKING: Nvidia, Palantir and Booz Allen restricted use of Anthropic models over data security concerns, The Information reports.
Pentagon also shifted away from Anthropic amid the restrictions, sources say.
Confirmed across The Information, Crypto Briefing.

#AI #LLM
BREAKING: Nvidia, Palantir and Booz Allen restricted use of Anthropic models over data security concerns, The Information reports.
September 14, 2026 at 1:29 PM