Liquid AI caught its model cheating on coding tests — then it recognized the sandbox
Liquid AI’s new reasoning model was built to solve coding problems, but during training it kept combing Git history, GitHub, The post Liquid A…
Open Source
⚑ Сообщить о проблемеПоследние новости Open Source от Tech News
Все темы
AI
agents
ai
api
architecture
automation
backend
beginners
career
database
devchallenge
devops
discuss
javascript
linux
llm
machinelearning
mcp
opensource
performance
productivity
programming
python
react
security
showdev
testing
tutorial
typescript
webdev
“Developers are paying frontier prices for amnesia”: Uvi wants to stop the agent relearning brain drain
Every time an AI agent has to trawl Slack threads, Jira tickets and old pull requests to piece together something The post “Developers are paying fr…
AWS, Upstage and Ollama agree on a decision-model API. OpenAI hasn’t signed on.
AWS’s new decision model comes with a familiar address. Released October 1, Strands Decider 2B is an open-weight model built The post AWS, Upstage a…
“Don’t use ‘open weight’ and ‘open source’ interchangeably”: Percona CEO on why AI terminology matters
The AI industry has enthusiastically embraced the language of open source, even as some of its most prominent “open” models The post “Do…
Your phone’s vector index might be bigger than the AI model running it
Google released EmbeddingGemma 2 on Tuesday, putting text, code, image, video, and audio retrieval into a 740-million-parameter open model that The…
Mistral’s new AI tried to escape its test environment. In three weeks, anyone can download it
Mistral launched Large 4 on Tuesday, its first major model release since Medium 3.5 at the end of April and The post Mistral’s new AI tried to escap…
The AI safety check that runs on a laptop and nearly matched a 35B model
Guardrail selection has typically meant choosing between a purpose-built classifier and an LLM acting as a judge. Decision models such The post The…
Eclipse wants companies to be free to switch AI providers. Today, doing so can mean a costly rebuild.
If you rely too heavily on one AI provider, switching can mean rebuilding workloads and moving data while you face The post Eclipse wants companies…
Claude Opus 5.5 vs. Fable 5.1: One overthinks, the other cuts corners
Anthropic released Claude Opus 5.5 on September 22, saying it performs at the level of Claude Fable 5.1 on most The post Claude Opus 5.5 vs. Fable 5…
Ember-1 vs. Kimi K3: Nearly identical results at 3.4 times the speed
Fireworks Research launched Ember-1 on September 23 as a research preview. Built on Moonshot’s open-weight model, Kimi K3, it claims The post Ember-…
You picked Claude Sonnet 5.5 — but Anthropic may send your request to Sonnet 5
Anthropic’s Claude Sonnet 5.5, released Monday, is the first Sonnet model to launch with cyber safeguards and model fallbacks like The post Yo…
Claude Opus 5.5 vs. Opus 5 on reasoning tasks: Cheaper, faster, but not better
When Anthropic released Claude Opus 5.5 this week, the company claimed the new model costs 40% less than Opus 5 The post Claude Opus 5.5 vs. Opus 5…
OpenAI makes you call sales for a custom voice. Google just made it self-serve.
Google released Gemini 3.8 Flash TTS and Gemini 3.8 Flash-Lite TTS today through the Gemini API and Google AI Studio. The post OpenAI makes you call…
“Impressive level of openness”: Xiaomi goes way beyond the usual open-weight playbook with MiMo-V2.6
New models are coming out thick and fast, almost on a weekly cadence, ranging from the powerful proprietary systems coming The post “Impressiv…
GPT-6 Sol closes most of the alignment gap with Astra. It’s one-fifth the price.
On Tuesday, OpenAI released GPT-6 Sol and Luna, an expansion of the GPT-6 line-up that aims to make GPT-6 Astra’s The post GPT-6 Sol closes mo…
OpenAI releases GPT-6 Sol and Luna — and cuts token prices in half
OpenAI on Tuesday released GPT-6 Sol and Luna, which will complement the flagship GPT-6 Astra model in OpenAI’s lineup. As The post OpenAI rel…
Anthropic releases Opus 5.5 and cuts pricing by 20%. Your agent calls might secretly get routed to an older model.
Claude Opus 5.5 is here, and Anthropic has lowered the price. The new model, released on Tuesday, costs $4 per The post Anthropic releases Opus 5.5…
Claude couldn’t hack OpenAI. Then Anthropic shipped Opus 5.
Three security researchers at Hacktron AI found a memory-corruption bug in a widely used image library. Finding it was the The post Claude couldn…
Intel squeezed a 1.58-bit LLM down to 1.485 bits without changing a single weight
The 1.58 in a 1.58-bit language model sounds like a hard limit, but Intel researchers pushed a ternary model below The post Intel squeezed a 1.58-bi…
“Machine translation is still broken for most of the world’s languages”: Cohere builds non-reasoning for a reason
Enterprise AI company Cohere announced North Small Translate last week, a mixture-of-experts (MOE) open-weight machine translation model that works…
Cohere says AI sovereignty matters — except it just gated its new translation model for commercial use
This week, Cohere released North Small Translate 1.0, but enterprises that want to deploy its open weights in production won’t The post Cohere…
“It could kill us all”: what Anthropic’s own researchers really think about superintelligence
On Tuesday evening, Anthropic pretraining researcher Jacob Coxon announced on X that he’d resigned. Within hours, two of his colleagues The po…
K2 Horizon just shipped as six new fully open models — developers aren’t fully convinced
Based in the Emirati capital, Abu Dhabi, the Institute of Foundation Models (IFM) introduced K2 Horizon last week. This group The post K2 Horizon ju…
GPT-6 Astra aced the hardest AI benchmark. The asterisk matters more than the score.
There was no mistaking the divide in March with the release of ARC-AGI-3. While frontier AI models could do little The post GPT-6 Astra aced the har…
“Hugging Face will remain an open platform”: Nvidia strikes $12.9B deal for the ‘GitHub of AI’
Nvidia has confirmed that it’s agreed to acquire Hugging Face in a mammoth $12.9 billion deal that will bring one The post “Hugging Face…
Nvidia PAIR lets you put your idle Macs and PCs to work for AI agents
Nvidia’s bet on open models and local AI has been taking shape for a while now. Its acquisition of Hugging The post Nvidia PAIR lets you put y…
Google ships its third Gemini Flash model in six weeks
Google hasn’t released a Gemini Pro model in a while, but on Wednesday, the company launched yet another set of The post Google ships its thir…
DeepSeek’s first vision model vs. Gemini 3.7 Flash: It comes down to spend vs. speed
DeepSeek released V4 Flash Vision Exp on August 21, its first model that accepts image input. Image input means a The post DeepSeek’s first vi…
Alibaba just released Qwen3.8-Flash: “An early preview of the architecture in Qwen4”
Alibaba this week unveiled Qwen3.8-Flash, an open-weight, multimodal Mixture-of-Experts (MoE) model.  Hot on the heels of Qwen 3.8 Max, which T…
Claude Opus 5 scored 30% on ARC-AGI-3. Wrapped in Nvidia’s AVO, it hit 100%.
Nvidia first introduced its Agentic Variation Operators (AVO) general-purpose coding agent system in late March 2026. The company has now The post C…