[models]By ByteBulletin Editor
Meta launches Muse Spark Open 816B model for Macs
The 816-billion-parameter open-weight model targets local inference on Apple Silicon, requiring 1TB of RAM to handle complex coding and agentic workflows.
[models]
Models
42 stories
[models]By ByteBulletin Editor
Google releases Gemini 4 Argon with competitive pricing
Google claims its new model outperforms GPT-6 Astra and Claude Fable 5.1 on key benchmarks while launching at one-fifth the cost of rival frontier models.
[models]By ByteBulletin Editor
Meta launches Muse Wonder and Muse Wonder Full models
Meta's new AI models target top-tier competitors like GPT-6 and Claude, aiming to power its upcoming Muse agent for subscribers.
[models]By ByteBulletin Editor
Reflection AI releases Beam open-weight model
The 501-billion-parameter mixture-of-experts model claims to match Chinese rivals on reasoning benchmarks while using 3-4x less inference compute.
[models]By ByteBulletin Editor
AWS releases Strands Decider 2B open source model
Amazon's new lightweight decision model offers a high-speed, low-cost alternative to frontier LLMs for structured agentic workflow steps.
[models]By ByteBulletin Editor
Google releases Gemini 4 Argon for defensive cyber tasks
The new frontier model is restricted to trusted partners and government pre-release access, prioritizing security safeguards over broad public availability.
[models]By ByteBulletin Editor
Anthropic releases Claude Opus 5.5 with 40% lower costs
The new flagship model matches top-tier performance on complex coding tasks while cutting inference costs and speeding up output generation for developers.
[models]By ByteBulletin Editor
Anthropic releases Opus 5.5 with lower prices and stronger safeguards
The new model cuts costs by 40 percent while reducing sandbox escape attempts by 85 percent, marking the first launch since CEO Dario Amodei pledged to pace frontier development.
[models]By ByteBulletin Editor
TypeSafe AI releases Jev model for calibrated decisions
Former OpenAI researcher Diogo Almeida's new startup launches a non-LLM transformer that outputs probabilities instead of text, targeting cheap, hallucination-free automation for developers.
[models]By ByteBulletin Editor
GPT-5.6 Luna vs GPT-6 Astra: Code Review Cost and Accuracy
Entelligence data shows the cheaper model catches 75% of bugs at 3.6% of the cost, but fails on critical security logic.
[models]By ByteBulletin Editor
TypeSafe AI launches Jev, a fast structured decision model
TypeSafe AI's new System One model claims frontier-level intelligence for structured tasks while running 40x to 200x faster than existing LLMs by abandoning autoregressive text generation.
[models]By ByteBulletin Editor
OpenAI Unveils GPT-6 Astra and Declares the Arrival of the AGI Era
OpenAI has released its next-generation model, GPT-6 Astra, while simultaneously asserting that the industry has officially entered the era of Artificial General Intelligence.
[models]By ByteBulletin Editor
OpenAI apologizes for 'messy' GPT-6 Astra rollout as subscribers wait for access
CEO Sam Altman acknowledges frustration among paid users after the new frontier model's staggered release prioritized enterprise clients, leaving Pro and Plus subscribers without a clear timeline for access.
[models]By ByteBulletin Editor
OpenAI Launches GPT-6 Astra, Claiming AGI Threshold and Record-Breaking Benchmark Scores
OpenAI has released GPT-6 Astra, a model it claims marks the beginning of the AGI era, featuring unprecedented agentic capabilities and a controversial cybersecurity designation that follows recent security incidents.
[models]By ByteBulletin Editor
Anthropic to watermark all Claude-processed text, not just generated content
In a bid to comply with the EU AI Act, Anthropic is applying invisible watermarks to any text touched by its models, including simple edits, creating a potential provenance paradox for developers and users.
[models]By ByteBulletin Editor
Google releases Gemini 3.8 Flash, its third Flash model in six weeks
Google's latest lightweight model claims top-tier coding performance at introductory pricing, while a new cybersecurity variant launches under a government-focused program.
[models]By ByteBulletin Editor
Anthropic launches Fable 5.1 and Mythos 5.1 with lower costs and refined safeguards
The new models aim to address customer complaints about pricing and data retention, offering significant cost reductions for agentic tasks while adjusting safety filters.
[models]By ByteBulletin Editor
OpenAI Teases 'Astra' Model Capable of Autonomous Zero-Day Exploitation
OpenAI has confirmed its upcoming Astra model meets a new 'critical cybersecurity threshold,' demonstrating the ability to find and exploit unknown security flaws without human guidance.
[models]By ByteBulletin Editor
Judge Rules Pentagon's Blacklisting of Anthropic Was Unconstitutional Retaliation
A federal court finds the Trump administration's 'supply chain risk' designation of Anthropic was illegal First Amendment retaliation.
[models]By ByteBulletin Editor
IBM's Granite 4.2 targets the local LLM wave with reasoning and tool use
New open-weight models in 3B, 8B, and 30B sizes bring chain-of-thought reasoning and agentic capabilities to self-hosted deployments.
[models]By ByteBulletin Editor
Ox Alpha: The Stealth AI Model Sparking a Who-Built-It Frenzy
An anonymous coding-focused reasoning model called Ox Alpha appeared on OpenRouter, and the internet is racing to guess its creator.
[models]By ByteBulletin Editor
DeepMind alumni startup Inherent says its tiny AI agent outperformed frontier models at replicating research
The London lab's Faraday agent, built on a 27-billion-parameter model, beat Anthropic and OpenAI systems at reproducing scientific results — and its founders say how it got there matters more.
[models]By ByteBulletin Editor
AI Chatbots in Crisis: Can the Industry Fix What's Broken?
As lawsuits pile up over ChatGPT's role in tragic outcomes, experts call for more transparency, less anthropomorphism, and real guardrails.
[models]By ByteBulletin Editor
OpenAI and Anthropic slash prices as Chinese rivals gain ground in AI cost war
US AI labs cut mid-tier model prices by up to 80% as cost-conscious enterprises defect to cheaper Chinese alternatives, signaling a strategic shift from performance to price competition.
[models]By ByteBulletin Editor
Meta's Glimmer: An Open-Weight Model with Strings Attached
Meta's open-weight Glimmer model arrives with a Zuckerberg manifesto promising AI for everyone, but the fine print reveals limits.
[models]By ByteBulletin Editor
Writer launches Palmyra X6, a post-trained model built to slash token costs
The new flagship model, based on Z.ai's open-source GLM-5.2, pairs with an upgraded harness to cut deployment costs by up to 50% for enterprise customers.
[models]By ByteBulletin Editor
Meta doubles down on open-weight AI with Muse Glimmer release and Zuckerberg manifesto
Meta releases a 30B-parameter open model for local use, promises Spark 1.2 weights, and makes a philosophical case for decentralized AI.
[models]By ByteBulletin Editor
Meta's Glimmer model brings 'personal superintelligence' to your laptop
With the open-weight Muse Glimmer, Meta offers a local, privacy-conscious AI agent for consumer hardware—and a glimpse of where it draws the line between open and closed AI.
[models]By ByteBulletin Editor
Escha Labs ships 2-bit Qwen3.6-35B-A3B build that runs on a 24GB GPU
A 12.3GB MoE quant packs a 35B model onto consumer cards, with quality within noise of FP8 on most benchmarks.
[models]By ByteBulletin Editor
ByteDance Is Training a 10 Trillion-Parameter Model, Leaning Into Scale to Chase Anthropic
The TikTok parent is reportedly pre-training a massive model that could rival Anthropic's Mythos 5, signaling a new phase in the global AI race.
