Skip to content

Search stories, models, companies…Subscribe

Topic · Concept

Frontier models

"Frontier models" are the most capable general-purpose AI models available at a given time, such as the top tiers from Anthropic, OpenAI and Google. The term is used in both industry and regulation, and its meaning moves as the state of the art advances.

73 stories · updated Oct 18

Follow via RSS Get Frontier models alerts

Key facts

Checked Oct 9

Frontier Model Forum: 501(c)(6) nonprofit, established 2023
FMF definition: General-purpose; outperforms models in wide use ≥12 months on benchmarks or high-risk capability assessments
EU AI Act threshold: > 10^25 FLOPs training compute presumes high-impact capabilities (Art. 51(2))
Example: Anthropic top tier: Claude Fable 5.1 ($10/$50 per MTok), Claude Opus 5.5 ($4/$20)
Example: OpenAI: GPT-6 Astra ($10/$50 per MTok in the API); GPT-6 Sol and Luna in ChatGPT since Oct 7, 2026
Example: Google: Gemini 4 Argon, announced Sep 30, 2026; limited access via the Fairwind Program; not yet on the Gemini API models page as of Oct 8, 2026

Latest on Frontier models

Agents · NewsThe Week in AI #4: Gemini 4 Argon ships, regulators move on rogue AI agents, and a wave of open weightsOctober 12–18, 2026: Gemini 4 Argon goes generally available, the FTC and India's CERT-In act on AI agent incidents, Anthropic's $16.8 billion quarter, and open-weight releases from Meta, Nvidia, Reflection and DeepSeek.Oct 18·7 min read

Open source · NewsReflection Beam open weights are out: 501B model on Hugging Face under Apache 2.0, in BF16, FP8 and NVFP4Reflection AI released the Beam open weights on October 16, 2026: a 501B-parameter (23B active) text model on Hugging Face under Apache 2.0, in BF16, FP8 and NVFP4 versions. Here's where to get Reflection Beam, what the model card's revised numbers say, and where the Nvidia talks stand.Oct 18·5 min read

Research · AnalysisMercor's benchmark map: why it says 90% of US work hours aren't covered by AI benchmarks, and what the number measuresMercor's benchmark map estimates that 90% of US working hours go to tasks with little or no coverage from AI benchmarks. Here's how Mercor got the 90% figure from O*NET tasks, 29 benchmarks and an LLM judge, what 'covered' means, and why Mercor's own business is part of the story.Oct 18·6 min read

Models · ExplainerWhy is Claude Fable 5 unavailable? The June export-control ban, its return, and what you can use nowWhy is Claude Fable 5 unavailable? The US government's June 12, 2026 export-control order forced Anthropic to switch it off worldwide. The ban was lifted on June 30 and Fable 5 came back on July 1. If you still can't pick it today, the cause is almost always your plan or usage credits.Oct 16·9 min read

Industry · ExplainerAnthropic IPO explained: what's confirmed about the date, price, ticker and exchange, and how shares are allocatedThe Anthropic IPO date, price and ticker haven't been announced. Anthropic has confirmed only a confidential draft S-1, filed June 1, 2026. Here's what's confirmed, what Bloomberg and Reuters have reported, and how IPO shares are generally allocated.Oct 16·8 min read

Research · Analysis"AI models can provably hide arbitrary capabilities": what the paper proves, and what it means for AI safety evaluationsAn October 2026 paper by researchers from Oxford and others shows how to lock a hidden skill inside a model's weights with cryptography, so that auditors with full weight access can't trigger or identify it. What 'provably' means here, the assumptions behind it, and why it matters for system cards.Oct 16·7 min read

Models · NewsGLM-Legal: what Z.ai's legal AI model is, how it scored, and whether you can download itZ.ai announced GLM-Legal on October 12, 2026, a legal-work model built on its open-weight GLM-5.3-Flash. Independent evaluator Vals AI ranks it seventh on Harvey's Legal Agent Benchmark at $0.15 a task, but the model itself is an application-only API beta with no published price, license or weights.Oct 16·5 min read

Policy · NewsFTC sends civil investigative demands to OpenAI and Anthropic over AI safety incidents: what a CID is and what happens nextOn October 14, 2026, the FTC sent civil investigative demands to OpenAI and Anthropic, each more than 50 pages and seeking records on security incidents, safeguards, marketing and privacy, Axios reported, citing one source. The FTC has confirmed only that the demands exist. Here's what the FTC civil investigative demands can compel, what they don't prove, and what comes next.Oct 16·6 min read

Open source · NewsMeta Muse Spark Open: the 816B open-weight model, its Muse Model License and what it takes to runMeta released Muse Spark Open on October 15, 2026: an 816B-parameter (42B active) open-weight model with a 1M-token context, under a new Muse Model License. What the license actually allows, what hardware it needs, and how Meta's own benchmark table stacks up.Oct 16·6 min read

Models · NewsMicrosoft MAI-Base-2: a 2.8-trillion-parameter base model, a 4–12x efficiency claim, and no release dateMicrosoft AI finished pretraining MAI-Base-2, a 2.8T-parameter mixture-of-experts model with 80B active parameters, on 16,000 GB200 GPUs, it said on October 15, 2026. What Microsoft claims, what Mustafa Suleyman reportedly told staff about OpenAI, and why you can't use it yet.Oct 16·4 min read

Nothing in this format on this page — try another filter or the next page.