The news desk
AI news, checked against the source.
What changed in AI, one story at a time. Each one links to its primary source, and facts stay apart from interpretation.
Last updated ·
Unlock 30 daysWednesday, November 18UTC
- NewGoogleProject Genie / Genie 4Pricing & Access
Google opens Genie 4 world creation to eligible AI Ultra subscribers
Google is rolling Genie 4 into Project Genie for eligible Google AI Ultra subscribers, with real-time world edits, native audio, a building agent and a community gallery. Google AI Pro members can join a waitlist. This is broader access to the model previewed for trusted testers in October, not a newly trained model.
Google2 sources

Free, email only
Unlock the full brief free.
- Thirty days of stories to browse, not seven
- On every story: what to check, each source and why it matters
- The same signup unlocks every free tool
Continuing subscribes the email on that account to Work Worth Doing, my newsletter, the same as the form below, and unlocks every free tool in this browser. We only receive your email address. Unsubscribe in one click.
Tuesday, November 17UTC
- NewElevenLabsScribe Speech to TextPricing & Access
ElevenLabs lowers Speech to Text to 20¢ batch and 22¢ realtime per hour
All Scribe models now cost $0.20 per audio hour for batch transcription and $0.22 for realtime on self-serve and Business plans. Enterprise pricing is unchanged.
ElevenLabs1 source
- NewOpenAIPro 200 GPT-6.2 Sol UltrafastPricing & Access
OpenAI adds GPT-6.2 Sol Ultrafast to Pro 200 Codex
Pro 200 subscribers can use GPT-6.2 Sol Ultrafast in Codex under their existing plan limits. Included use draws down at 6× Standard; purchased credits use a promotional 3× rate through February 28, 2027. Astra Ultrafast remains unavailable on Pro 200.
OpenAI2 sources
- NewOpenAIGPT-6.2 Astra API pricingPricing & Access
OpenAI cuts Astra API token rates with GPT-6.2
GPT-6.2 Astra API Standard pricing is $8 input, $0.40 cached input and $40 output per million tokens. OpenAI says input and output rates are 20% below GPT-6 Astra and cached input is 60% lower; GPT-6.2 Sol keeps the prior Sol rate.
OpenAI1 source
- NewElevenLabsElevenAgents model lifecyclePricing & Access
ElevenAgents schedules GPT-4 family retirement and fallback migration
ElevenLabs deprecated GPT-4.1, GPT-4.1-mini, GPT-4o and GPT-4o-mini in ElevenAgents before a January 14, 2027 shutdown. Requests start a 10% GPT-6 Luna fallback December 15, then move fully January 9. GPT-6 Luna is made by OpenAI, not ElevenLabs.
ElevenLabs1 source
- NewMicrosoftMAI Licensed WeightsPricing & Access
Microsoft announces licensed MAI weights for customer Azure environments
Microsoft says selected organizations will be able to customize and deploy its MAI model weights in their own Azure environments, controlling access and upgrades. The offering is coming soon through Frontier Company engagements; terms and public availability were not announced. These are licensed weights for selected customers, not a general open-weight release.
Microsoft AI1 source
- NewOpenAIResponses API background completion windowsPricing & Access
OpenAI adds discounted two and eight hour API completion windows
The Responses API public beta now offers two hour and eight hour background completion windows at 30% and 40% below Standard model token rates. It supports GPT-6.2 Astra and Sol, GPT-6.1 Terra and GPT-6 Luna; discounted work shares Batch/Background limits.
OpenAI2 sources
- NewSunoVisuals creditsPricing & Access
Suno sells Visuals credits separately from song credits
Suno says Visuals uses credits purchased separately from song creation credits. Its public purchase page requires account context and did not expose a trustworthy fixed price in the available read, so no per-video rate is asserted.
Suno2 sources
- NewOpenAIGPT-6.2 on Amazon BedrockPricing & Access
AWS makes OpenAI GPT-6.2 Astra and Sol available on Bedrock
AWS announced general availability of OpenAI-made GPT-6.2 Astra and Sol on Amazon Bedrock, with supported regions, endpoints, model tiers and prices in AWS documentation. AWS hosts the models; it did not train them.
AWS2 sources
- NewKlingKling Image 4.0 API billingPricing & Access
Kling changes Image 4.0 API billing to include reference images
Kling’s Image 4.0 API update says billing counts both input reference images and output images, rather than outputs only, with output rates varying by resolution. Exact current rates were not accessible in the public read, so users should price a representative job before scaling.
Kling1 source
Monday, November 16UTC
- NewElevenLabsElevenCreative annual plansPricing & Access
ElevenLabs offers temporary annual plan discounts for creators
ElevenLabs announced a limited ElevenCreative annual-plan promotion from November 16 through December 1: eligible Creator annual plans are $99 for the first year, new Pro annual subscribers get 55% off and monthly Pro subscribers moving to annual get 50% off. Eligibility and renewal terms matter.
ElevenLabs2 sources
Friday, November 13UTC
- NewDeepSeekDeepSeek file and voice API pricingPricing & Access
DeepSeek sets separate file and voice API usage rates
DeepSeek’s newly opened file parsing and OCR APIs cost $0.225 per 1,000 pages. Text to speech lists $0.45 per 10,000 characters and speech recognition $0.116 per hour, each temporarily half price at $0.225 and $0.058 respectively. These fees add to model-token charges; DeepSeek gives no verified end date for the promotion. Search and Fetch rates were already announced in October.
DeepSeek2 sources
- NewGoogleGoogle AI Studio creditsPricing & Access
Google adds in-Studio claiming for eligible monthly cloud credits
Google AI Pro and Ultra subscribers can now claim existing Google Developer Program monthly Cloud credits inside AI Studio and connect billing there. The credits are tied to eligibility and an active Developer Program profile; Google did not announce a new universal free-credit amount.
Google1 source
Days are UTC. Some sources report an exact time and others only a date; both are placed on their UTC day.

