News✳
The releases, research and policy decisions shaping AI around the world. Original context, with sources you can follow.
Read the announcement. Understand the implications.
Coverage distinguishes dated events, company-reported claims and editorial analysis. Follow a topic guide to understand the concepts behind a release, infrastructure decision or policy development.

Gemini Tool Retrieval enters preview: smaller prompts, on-demand tools and the cost tradeoff
Google’s new Interactions API feature loads tool parameter schemas only when needed. It can reduce a large agent’s input burden, but the benefit depends on which tools it actually uses.

AssemblyAI STT 3.7 arrives with three models: prices, speaker labels and migration dates
The new speech-to-text family separates recordings, live captions and voice agents. Existing users should also check the December default-model switch and changes to transcript output.

Tasks in Gemini goes free in India: what the rollout includes and what stays U.S.-only
Google is opening its background task agent to six countries. India gets the free rollout, while checkout, texting and email entry points begin with narrower U.S. eligibility.

Qwen4-Max leaves preview: API access, regional pricing and the limits to check
Alibaba’s November 16 release adds a production route beyond the restricted preview. Regional prices, supported features and a separate preview retirement date deserve attention before switching.

Gemini 4 Hydrogen explained: pricing, free access and what changes in Google Search
Google’s fast execution model is now part of coding tools, the Gemini app and Search. Here is what it does, what access costs and which limits matter before you use it.

Suno Bridge brings AI audio into Ableton Live: access, credits and the upload to know
Suno’s new Mac beta generates instrument and vocal parts inside an existing production workflow. It requires Premier, uses credits and uploads the selected project to understand its musical context.

ChatGPT Wallet launches: who can use it and what purchase approval means
OpenAI’s new payment feature lets eligible users authorize purchases through dots. The launch is US-only, and approving a payment is still different from receiving a confirmed order.

DeepSeek opens its app toolkit: files, links and voice for custom chatbots
DeepSeek’s November 13 release brings more of its chat experience to developers. File parsing choices, separate tool charges and response handling will shape what a useful integration looks like.

Gemini Decisions API: Google adds fast scoring for choices, routing and checks
Google’s November 11 release scores fixed options across multimodal inputs. Its documentation makes the boundaries clear: independent questions, a shared context budget and application code that decides what happens next.

OpenAI brings businesses into ChatGPT: Business Hub, agents and what changes
Businesses can manage their presence inside ChatGPT and, where eligible, launch a customer-facing agent. Here is what the beta offers, how access differs, and what it means for Indian businesses.

Meta’s Muse Ember targets high-volume agents, with a long-context pricing catch
Meta’s lightweight model is available for repeatable coding and document work. Its low headline rates rise for prompts above 200,000 tokens, making task design part of the cost calculation.

Claude Fable 5.2 and Mythos 5.2: Pricing, Access and API Migration Changes
Anthropic’s November 9 release pairs broader Fable access with verified access to Mythos. For developers, effort settings and conversation handling matter alongside the new token prices.