Alibaba discounts its 2.4T Qwen preview first
Qwen3.8-Max-Preview arrives with 2.4T parameters, Qoder access and steep credit discounts before a full model card or independent benchmarks.
AI models News and Analysis
14 verified stories covering AI models, product updates and industry developments.
Google turns Flash models toward cost
Google released three Gemini Flash models at once, splitting general, low-cost and cybersecurity workloads into separate products.
Grok 4.5 targets coding seats with low prices
Grok 4.5 targets coding seats with low prices. A concise localization of the verified facts and the industry signal behind the story.
Meta prices Muse Spark 1.1 for agent work
Meta prices Muse Spark 1.1 for agent work is reframed for global technology readers, with the key numbers, product claims and open questions kept intact.
Ollama raises $65 million for local models
Ollama raises $65 million for local models is reframed for global technology readers, with the key numbers, product claims and open questions kept intact.
Microsoft routes some Office AI prompts to MAI models
Microsoft routes some Office AI prompts to MAI models is reframed for global technology readers, with the key numbers, product claims and open questions kept intact.
MoWorld runs a 50 FPS world model on Chinese NPUs
MoWorld runs a 50 FPS world model on Chinese NPUs is reframed for global technology readers, preserving the key numbers, claims and open questions.
Google I/O 2026: Gemini 4.0 Isn't the Headline Anymore
Google I/O 2026 kicks off tomorrow, but Gemini's big update is no longer the main event as the company shifts focus to AI agents and hardware.
Grok 4.3 Beta Launches: TTS 90% Cheaper Than Competitors
xAI released Grok 4.3 Beta on April 17, adding native video analysis, document generation, and developer voice APIs priced 86-92% below rivals. The $300/month tier still lacks conversation memory.
MiniMax M2.7 Self-Iterates 100 Rounds, Performance Up 30%
MiniMax's M2.7 model used an agent framework to automatically run over 100 training iterations without human intervention, boosting internal benchmarks by 30%. The 230B-parameter sparse MoE model activates only 10B per inference and is now open-source.
Google launches its cheapest Gemini model: Flash-Lite
Google has released Gemini 3.1 Flash-Lite, its most affordable AI model to date, priced at $0.25 per million input tokens and $1.50 per million output tokens. The model offers faster generation speeds and a 1M-token context window, targeting high-volume, cost-sensitive applications.
Meta Launches Its First Closed-Source Model, Muse Spark
Meta released Muse Spark, its first fully proprietary flagship AI model, marking a strategic shift from open-source Llama toward competing directly with OpenAI and Google.
Microsoft Develops Three In-House AI Models to Reduce Reliance on OpenAI
Microsoft quietly launched three proprietary AI models on April 2, aiming to lessen its dependence on OpenAI. The models include speech-to-text, text-to-speech, and image generation tools, with a strategic focus on distribution and pricing leverage.
Claude Mythos tops SWE-bench Pro at 77.8%
Anthropic unveiled Claude Mythos, its most powerful model yet, but is only granting preview access to 40 organizations under Project Glasswing for defensive cybersecurity work.