Databricks raised $3 billion at a $188 billion valuation, a 40% jump from its $134 billion mark five months ago, and revealed that its 3,000 engineers are defaulting to Chinese open-source model GLM 5…
Databricks announced a new funding round led by Coatue that values the company at $188 billion, extending its fundraising streak as it transforms into an AI provider. The round is expected to close la…
Chinese leader Xi Jinping called for more open-source AI and global regulation at the World Artificial Intelligence Conference in Shanghai on Friday, stating 'China is ready to be more open.' Xi outli…
Kilo users can now match skills to models by building custom agents that bundle a model, system prompt, and tool permissions into a single profile, enabling one-keystroke switching between agents like…
GLM 5.2 is 35% off via Novita on AI Gateway through July 24. To get the discounted rate, set the model to 'zai/glm-5.2' in the AI SDK and route requests through Novita using gateway provider options. …
LM Studio launched LM Studio Bionic, an AI agent for open models that supports coding, research, and document work with local or cloud-based open-source models, committing to zero data retention and n…
LM Studio launched LM Studio Bionic, a new Mac app that uses open models for coding, research, and document work, with support for local and cloud-based open-source models under a Zero Data Retention …
Kimi's new K3 model, with 2.8 trillion parameters and a one-million-token context window, challenges industry leaders like GPT 5.6 Sol and Claude Fable 5, but its significantly higher cost signals a p…
Kimi is launching K3, a multimodal open-weight model with 2.8 trillion parameters and one million tokens of context, which in its own benchmarks approaches GPT-5.6 Sol and Claude Fable 5 while surpass…
A researcher demonstrated that an open-weight coding model can be turned into a backdoor in under one hour and for less than $100, raising concerns about trust in Chinese open weight models like GLM 5…
Hugging Face disclosed a security incident in July 2026 where an autonomous agent framework gained unauthorized access to internal datasets and credentials by exploiting code-execution paths in its da…
OpenAI's GPT-5.6 Sol ranks first overall on Design Arena's Web Design (Non-Agentic) Arena, 18 places higher than its predecessor GPT-5.5, marking the first time an OpenAI model has topped the leaderbo…
Dax Raad, co-founder of OpenCode, revealed on the Syntax podcast that his team's AI token usage increased 5x after gaining early access to an unreleased AI model, and when access ended, team members m…
Zhipu AI's open-weight GLM 5.2 model matches or outperforms closed models 10 to 20 times larger on function calling and structured design benchmarks, signaling a shift in the viability of smaller open…
Databricks reveals that AI token cost does not always equate to task efficiency, with its internal coding benchmark showing that models with cheaper per-token costs can incur higher overall task costs…
Tracebit researchers found that planting prompt injections alongside secrets on AWS reduced AI hacking agents' success rate from 57% to 5% for seizing admin access and from 36% to 1% for complete comp…
Open-weight models surged to 29% of AI Gateway token volume in June 2026, up from 11% in April, on under 4% of spend, according to Vercel's AI Gateway Production Index. DeepSeek reached 22.6% of token…
Z.AI's GLM 5.2, an open-weight coding model, offers developers a cost-effective alternative to existing tools by autonomously auditing codebases, redesigning UIs, and hunting bugs from Sentry and Verc…
A developer detailed the local deployment of GLM 5.2, a 753B MoE model with 40B active parameters requiring 1.51TB of RAM in BF16 precision. Unsloth's dynamic quantizations, such as UD-Q3_K_XL at 343G…
Perplexity released a research preview of a post-trained version of Z.AI's GLM 5.2, a Chinese open-source model, that matches Claude Opus 4.8 performance at one-third the cost by using an advisor tool…