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…
Latest News
⚑ Report a ProblemTech news from the best sources
All topics
- игры
AI
Gear
News
Tech
agents
ai
api
architecture
automation
beginners
career
database
devops
javascript
llm
machinelearning
mcp
opensource
performance
productivity
programming
python
react
security
showdev
testing
tutorial
typescript
webdev
Why basic RAG fails at multi-hop reasoning (and how GraphRAG fixes it)
The current approach to designing LLMs within AI engineering is oversimplified. According to the echo chamber’s view, solving LLM hallucinatio…
Claude Desktop can now easily run Qwen, DeepSeek and Kimi models — after Ollama’s first effort stalled
Open-weight model runner Ollama has reintroduced an integration with Claude Desktop that lets users connect Anthropic’s app to models served T…
“Save frontier models for frontier problems”: Why Korea’s Solar Pro 4 is a workhorse agent reliability play
South Korean AI model company Upstage AI officially announced the launch of its Solar Pro 4 closed commercial LLM last The post “Save frontier…
An industrial-scale distillation of models, or subtle benchmaxxing: What developers really think of GLM-5.3
Chinese frontier model outfit Z.ai released GLM-5.3 on Friday, a model hewn from the same codebase as its predecessor GLM-5.2, The post An industria…
A Claude Code skill was eating 200,000 tokens before answering a single question
A Claude Code skill designed to help developers work with Anthropic’s API was consuming more than 200,000 tokens to load. The post A Claude Co…
OpenAI’s Greg Brockman: Z.ai’s GLM-5.3 likely to “significantly accelerate the threat landscape”
OpenAI has adopted a less-than-straightforward stance with regards to open-weight AI models, both raising alarms over powerful Chinese releases whil…
Apple’s new AI split means your iOS app could behave differently in China
Apple is splitting up its AI stack. Instead of rolling out the same system worldwide, the company reportedly built a The post Apple’s new AI s…
GLM-5.3 didn’t change the base model — where did its coding gains come from?
Z.ai released GLM-5.3 on Friday, a coding and agent model built from the same base model as GLM-5.2. Developers can The post GLM-5.3 didn’t change t…
Why your AI pipeline costs 10x more after the demo
Every token has a price. The problem is that most AI systems don’t reveal the bill until they reach production. The post Why your AI pipeline costs…
Meta stopped worrying about distillation and just shipped the pipeline
Meta released Muse Glimmer on Monday, a 30-billion-parameter open-weight model distilled from Muse Spark and licensed under Apache 2.0. The The post…
OpenAI built a model it doesn’t want most people to use
OpenAI on Monday released GPT-5.6 Cyber, a model trained for security work that general-purpose models routinely block. It’s now available The…
“It blows my mind”-“It has a tendency to overengineer things a little”: Developers react to road-testing OpenAI GPT‑5.6 Sol
OpenAI made its family of GPT-5.6 models available to its app and API users globally at the start of July.  The post “It blows my mind”-“…
GPT-5.6 Sol just got better in one place and stayed the same everywhere else
Teams testing prompts in ChatGPT before moving them to Codex or Work may notice the difference on longer tasks. OpenAI The post GPT-5.6 Sol just got…
OpenAI’s Astra just proved 10 long-standing math and science theorems. The tokens cost $2,000.
OpenAI shared a big research update this week, announcing that an internal version of Astra, its next frontier model, created The post OpenAI’…
Alibaba Qwen3.8-Max reactions: “An API business model wearing an open source jacket”
Alibaba this week announced the launch of Qwen3.8-Max. The most powerful model in the Qwen series to date, this multimodal The post Alibaba Qwen3.8-…
Opus 5 vs. Fable 5: What does half the price buy?
Anthropic released Claude Opus 5 last week, with this pitch: It comes “close to the frontier intelligence of Claude Fable The post Opus 5 vs. Fable…
“This is not in my top ten list of worries”: What Sam Altman thinks about model distillation
Sam Altman’s latest appearance on Patrick O’Shaughnessy’s Invest Like the Best podcast covered everything from AGI and robotics to…
“Second only to Fable 5:” Alibaba talks the talk with Qwen3.8 without providing any real data
Alibaba has revealed Qwen 3.8, its latest, greatest large language model (LLM) which, it says, is among the most powerful The post “Second onl…
Claude Fable 5 vs. Kimi K3: Same results, one-third the cost, 4x slower
Moonshot AI released Kimi K3 in mid-July, selling it as a serious professional coding tool that competes head-on with Claude The post Claude Fable 5…
Kimi K3 tops Arena’s coding leaderboard — and it’s open-weight
Developers building with AI have largely relied on proprietary models like Anthropic’s Fable and OpenAI’s GPT-5.6 Sol for their most The…
Open-source AI is just “4 months behind” closed frontier models — and 10x cheaper
A quiet revolution is brewing in the AI model space. The dominance of proprietary closed frontier models has been cemented The post Open-source AI i…
AI can finally read your handwriting — here’s why enterprises care
The seemingly unquenchable thirst of the AI data ingestion pipeline spans language, numerical, and tabular data in the first instance, while Th…
Anthropic extends Fable 5 again — and won’t talk about what developers found inside Cursor
Anthropic has extended its enhanced access to Claude Fable 5 for all paid subscribers through July 19, marking the third The post Anthropic extends…
Meta debuts Muse Spark 1.1 and it isn’t free
Meta on Thursday rolled out Muse Spark 1.1, a major update to its AI platform, just three months after launching The post Meta debuts Muse Spark 1.1…
The “silent hallucination” loop: how our autonomous data pipeline poisoned its own vector store
Nothing matches the dread of checking a perfectly green observability dashboard with sub-100ms latency, right before an enterprise client emails The…
“Opus-class, but faster”: What Elon Musk says about beating Anthropic
SpaceXAI CEO Elon Musk announced on Wednesday that Grok 4.5 will be released publicly on Thursday, posting on X, “Based The post “Opus-c…
Meta says it caught OpenAI. One thing is missing.
The same week Mark Zuckerberg told Meta staff that the company’s AI bets “haven’t come to fruition yet,” his superintelligen…
OpenAI’s own safety card says GPT-5.6 has a lying problem
OpenAI’s confirmation late Tuesday that GPT-5.6 Sol, Terra, and Luna will launch publicly on Thursday closes one of the more The post OpenAI&#…
“Nature is the most computationally efficient system we know”: How Refiant used swarm optimization to build a 10-million-token AI model
While the household-name frontier models race forward with version numbers and context windows of at least a million tokens, a The post “Nature is t…