DeepSeek Rivals Rally as Open Source LLM War Heats Up
Global competitors including Meta, Mistral, and Alibaba accelerate open source model releases as DeepSeek's cost-efficie…
📂 LLM News
Latest articles in LLM News
Cohere Launches Command R Plus Fine-Tuning API
Cohere opens fine-tuning capabilities for its flagship Command R Plus model, targeting enterprise customers seeking cust…
Grok 3.5 Tops GPT-4o on MMLU Benchmark
Elon Musk's xAI releases Grok 3.5, claiming top scores on the MMLU benchmark and challenging OpenAI's dominance in LLM p…
Apple Intelligence 2.0 Brings On-Device LLM to macOS Tahoe
Apple's next-gen AI suite debuts a fully on-device large language model for macOS Tahoe, prioritizing privacy and perfor…
Meta Launches Llama 4 Maverick With 128K Context
Meta releases Llama 4 Maverick open-weight model featuring 128K context window, challenging proprietary rivals from Open…
Claude 4 Opus Enters Internal Testing Phase
Anthropic reportedly begins internal testing of Claude 4 Opus this summer, signaling a major leap in AI model capabiliti…
Krutrim Launches Open Source Hindi-First LLM
Indian AI startup Krutrim releases an open-source large language model built for Hindi and Indic languages, challenging …
South Korea ETRI Builds Lightweight Korean LLM for Gov Use
South Korea's ETRI develops a compact Korean language model optimized for government services, aiming to reduce reliance…
Preferred Networks Launches PLaMo 2.0 Open Source LLM
Japanese AI leader Preferred Networks releases PLaMo 2.0, an open-source foundation model designed to rival Western LLMs…
Small Language Models Beat Giants in Domain Tasks
Small language models with under 7B parameters are outperforming GPT-4 and other large models in specialized domain task…
Snowflake Arctic Embed 2.0 Redefines Text Embedding
Snowflake releases Arctic Embed 2.0 with multilingual, multimodal capabilities and top-tier benchmark scores across all …
Databricks DBRX 2.0 Targets Enterprise Analytics
Databricks unveils DBRX 2.0, a foundation model built specifically for enterprise data analytics workloads with major pe…