AllLLMCoding Co-pilotPersonal AssistantsAI AgentsHealthcare AIEthics & RegulationResearch & PapersStartups & Funding
Company:AllOpenAIAnthropicGoogleMicrosoftAppleDeepSeekMistralxAIMetaOthers

LLM · Other Companies

German AI consortium releases Soofi S, an open 30B model that tops benchmarks in both English and German
Featured
LLM

German AI consortium releases Soofi S, an open 30B model that tops benchmarks in both English and German

A German research consortium has released Soofi S 30B-A3B, an open language model trained entirely on Deutsche Telekom's cloud infrastructure in Munich. The model uses an efficient hybrid...

The Decoder
Read more
Mira Murati’s Thinking Machines Lab Makes The Technical Case For Human-Centered AI Built On Customizable Model Weights
LLM

Mira Murati’s Thinking Machines Lab Makes The Technical Case For Human-Centered AI Built On Customizable Model Weights

Mira Murati's Thinking Machines Lab published an essay on human-centered AI that frames human participation, model...

MarkTechPost
Open source AI matters more than ever, according to Hugging Face’s Clem Delangue
LLM

Open source AI matters more than ever, according to Hugging Face’s Clem Delangue

Hugging Face CEO Clem Delangue highlights the growing importance of open source AI, noting that the platform has become...

TechCrunch Startups
Hugging Face’s CEO on why companies are done renting their AI
LLM

Hugging Face’s CEO on why companies are done renting their AI

Hugging Face CEO Clem Delangue discusses the growing trend of open source AI adoption, with the platform now used by...

TechCrunch Startups
Meet Nemotron Labs 3 Puzzle 75B A9B: A Compressed Hybrid MoE LLM Delivering 2.03x Server Throughput
LLM

Meet Nemotron Labs 3 Puzzle 75B A9B: A Compressed Hybrid MoE LLM Delivering 2.03x Server Throughput

NVIDIA released Nemotron-Labs-3-Puzzle-75B-A9B, a compressed hybrid MoE language model that reduces parameters from...

MarkTechPost
Fast token generation emerges as the key differentiator as heterogeneous inference takes hold
LLM

Fast token generation emerges as the key differentiator as heterogeneous inference takes hold

The article discusses how fast token generation has become the critical differentiator in AI inference, moving from...

SiliconANGLE
NVIDIA Releases Nemotron-Labs-3-Puzzle-75B-A9B: A Compressed Hybrid MoE LLM Delivering 2.03x Server Throughput at Matched User Throughput
LLM

NVIDIA Releases Nemotron-Labs-3-Puzzle-75B-A9B: A Compressed Hybrid MoE LLM Delivering 2.03x Server Throughput at Matched User Throughput

NVIDIA released Nemotron-Labs-3-Puzzle-75B-A9B, a compressed hybrid MoE language model that reduces parameters from...

MarkTechPost
Datalab Lift vs the Field: How a 9B Schema-First Extractor Compares with NuExtract3, LlamaExtract, Marker, and Docling
LLM

Datalab Lift vs the Field: How a 9B Schema-First Extractor Compares with NuExtract3, LlamaExtract, Marker, and Docling

Datalab's Lift is a 9B schema-first document extraction tool that converts PDFs and images directly to JSON without...

MarkTechPost
I Built a Self-Improving AI, and So Can You
LLM

I Built a Self-Improving AI, and So Can You

The article discusses experiments where AI is used to build and improve other AI systems, demonstrating that advanced...

Wired AI
Chinese AI startup MiniMax plans to open-source a 2.7 trillion parameter model later this year
LLM

Chinese AI startup MiniMax plans to open-source a 2.7 trillion parameter model later this year

Chinese AI startup MiniMax is developing a large language model with 2.7 trillion parameters and plans to release it as...

The Decoder
NVIDIA Releases Audex (Nemotron-Labs-Audex-30B-A3B): A Unified Audio-Text LLM That Preserves the Text Intelligence of Its Backbone
LLM

NVIDIA Releases Audex (Nemotron-Labs-Audex-30B-A3B): A Unified Audio-Text LLM That Preserves the Text Intelligence of Its Backbone

NVIDIA released Audex (Nemotron-Labs-Audex-30B-A3B), a unified audio-text large language model that combines audio...

MarkTechPost
Cohere Transcribe Arabic is an open-source model built for Arabic's toughest transcription problems
LLM

Cohere Transcribe Arabic is an open-source model built for Arabic's toughest transcription problems

Cohere has released Transcribe Arabic, an open-source speech recognition model with 2 billion parameters that...

The Decoder
Liquid AI Open-Sources Antidoom: A Final Token Preference Optimization (FTPO) Method that Reduces Doom Loops in Reasoning Models
LLM

Liquid AI Open-Sources Antidoom: A Final Token Preference Optimization (FTPO) Method that Reduces Doom Loops in Reasoning Models

Liquid AI released Antidoom, an open-source method that eliminates doom loops in reasoning models by using Final Token...

MarkTechPost
Chinese AI models regularly pass 30 percent on OpenRouter as cost gap widens
LLM

Chinese AI models regularly pass 30 percent on OpenRouter as cost gap widens

Chinese AI models are increasingly being adopted by US companies on OpenRouter, with some models achieving over 30...

The Decoder
Tencent Releases Hy3: An Open 295B Mixture-of-Experts (MoE) Model with 21B Active Parameters and 256K Context
LLM

Tencent Releases Hy3: An Open 295B Mixture-of-Experts (MoE) Model with 21B Active Parameters and 256K Context

Tencent released Hy3, a 295B Mixture-of-Experts model with 21B active parameters and 256K context window, available...

MarkTechPost
Tencent releases Hy3 open-source model that allegedly matches models up to five times its active size
LLM

Tencent releases Hy3 open-source model that allegedly matches models up to five times its active size

Tencent released Hy3, an open-source language model with 295 billion total parameters but only 21 billion active...

The Decoder
GPT-4's dominance lasted a year while today's top models barely survive seven weeks at the top
LLM

GPT-4's dominance lasted a year while today's top models barely survive seven weeks at the top

GPT-4 dominated the Epoch Capabilities Index for approximately one year, but recent AI models have much shorter tenures...

The Decoder
Reddit is using LLMs to solve a problem LLMs largely created
LLM

Reddit is using LLMs to solve a problem LLMs largely created

Reddit is deploying Large Language Models to combat spam on its platform, addressing a problem that LLMs themselves...

TechCrunch AI
Nvidia's Kyber NVL144 reportedly pushed back more than a year, Asian suppliers drop
LLM

Nvidia's Kyber NVL144 reportedly pushed back more than a year, Asian suppliers drop

Nvidia's Kyber NVL144 AI server rack has been delayed over a year to 2028 due to circuit board manufacturing issues,...

The Decoder
Amazon sunsets Mechanical Turk, the original "Artificial Artificial Intelligence"
LLM

Amazon sunsets Mechanical Turk, the original "Artificial Artificial Intelligence"

Amazon Web Services is shutting down Mechanical Turk, its crowdsourcing platform that provided human workers for tasks,...

The Decoder

Stay Updated

Get the latest AI news delivered to your inbox every morning. No spam, unsubscribe anytime.