LLMAnthropic
An AI model programmed nonstop for 19 days on a single MirrorCode task that cost $2,600 to run
Epoch AI's MirrorCode benchmark tests whether AI models can recreate complete programs without access to original code....
The Decoder
LLM
Large Language Models
LLMOpenAI
OpenAI Has New AI Models. Here’s Why You Can’t Use Them
The White House requested OpenAI delay the rollout of its GPT-5.6 AI models, following a similar incident where...
Wired AI
LLMOpenAI
OpenAI unveils GPT-5.6 amid US AI regulatory drama
OpenAI has released GPT-5.6, a new model suite with three tiers (Sol, Terra, and Luna) designed for coding,...
The Verge AI
LLMOpenAI
Previewing GPT-5.6 Sol: a next-generation model
OpenAI announces GPT-5.6 Sol, a next-generation large language model with enhanced capabilities in coding, science, and...
OpenAI Blog
LLMAnthropic
Anthropic’s Claude is winning over paid consumers, a market owned by ChatGPT
Anthropic's Claude is gaining market share among paid AI consumers, challenging ChatGPT's dominance in the premium AI...
TechCrunch AI
LLM
DeepReinforce Releases Ornith-1.0: An Open-Source Coding Model Family That Learns Its Own RL Scaffolds
DeepReinforce released Ornith-1.0, an open-source coding model family built on Gemma 4 and Qwen 3.5 that learns its own...
MarkTechPost
LLM
Baidu Releases Unlimited OCR, a 3B Model That Keeps the KV Cache Flat for Long-Document Parsing
Baidu open-sourced Unlimited OCR, a 3B-parameter MoE model that efficiently parses long documents using Reference...
MarkTechPost
LLMOpenAI
OpenAI, Broadcom debut custom Jalapeño chip for AI inference
OpenAI has unveiled a custom chip called Jalapeño developed in collaboration with Broadcom to power its large language...
SiliconANGLE
LLMOpenAI
OpenAI and Broadcom Introduce AI Inference Chip
OpenAI and Broadcom have introduced a new AI inference chip designed to reduce token costs for AI model users. The chip...
AI Business
LLM
Companies are scrambling to stop employees from maxing out AI budgets with small tasks
Companies are implementing token rationing and budget controls to prevent employees from depleting AI API quotas with...
TechCrunch AI
LLM
Gradium Launches stt-translate and s2s-translate, Real-Time Speech Translation Models Beating gpt-realtime-translate on Accuracy and Latency
Gradium released stt-translate and s2s-translate, real-time speech translation models that support 20 language pairs...
MarkTechPost
LLMOpenAI
OpenAI says ChatGPT Instant now better understands what users actually want
OpenAI is updating GPT-5.5 Instant with improvements to conversation quality, including better intent recognition,...
The Decoder
LLM
Snowflake CEO finds GLM-5.2 competitive with Opus 4.7 at a fraction of the cost
Zhipu AI's GLM-5.2 model demonstrates competitive performance with Claude Opus 4.7 on coding benchmarks while costing...
The Decoder
LLMOpenAI
OpenAI reveals its first AI processor: Jalapeño
OpenAI has unveiled Jalapeño, a new AI processor chip developed in partnership with Broadcom designed specifically for...
The Verge AI
LLMMistral
Mistral AI Tackles Unstructured Data Challenge with OCR 4
Mistral AI has released OCR 4, a model designed to handle unstructured data with features like bounding boxes for...
AI Business
LLMOpenAI
OpenAI and Broadcom unveil "Jalapeño," a custom chip built for LLM inference
OpenAI and Broadcom have jointly developed the "Jalapeño" custom chip specifically designed for large language model...
The Decoder
LLM
Pangram CEO says language models give themselves away by making the same arguments
Pangram CEO Max Spero argues that language models can be identified by their homogeneous reasoning patterns, as they...
The Decoder
LLMOpenAI
Samsung opens ChatGPT Enterprise and Codex access after AI restrictions
Samsung Electronics is expanding employee access to ChatGPT Enterprise and Codex, enabling broader use of AI tools...
AI News
LLMMistral
Mistral's new OCR model beats competitors in 72 percent of blind test cases, company says
Mistral AI has released OCR 4, a new optical character recognition model designed to extract text from documents like...
The Decoder