219 summarised stories about Model Releases, each linking back to the original source. Browse all topics →
Sakana AI 12
Sakana AI launched the newsletter “Sakana AI Insider” to share its work from research through industrial implementation and product development.
Ben's Bites 1 week ago 43 ● 16 sources
Ben’s Bites reported a roundup of new AI agents, model releases, and tooling updates across companies including ChatGPT, Anthropic, Google, Grok, and various productivity apps. The most specific figure mentioned is that Claude Haiku 5.5 costs about 75% less to run than Haiku 4.5. As a result, more interfaces and image/chart/map-capable outputs are rolling out to users, while some services add usage credits or new integrations for building agent workflows in apps like Slack, Docs, and Cursor.
Trending Topics 1 week ago 44 ● 2 sources
Elon Musk announced that SpaceXAI’s Grok Bot will stop relying only on in-house Grok models and will use a different “best back end” model per task. Claude Opus 5.5, Midjourney, and Suno are named as examples of third-party models Grok Bot may route to. This shifts Grok Bot to model routing across multiple providers instead of using Grok exclusively, and it does not specify which tasks go to which model.
Trending Topics 1 week ago 32 ● 2 sources
Google will remove Gemini Flash and Gemini Pro from free personal accounts starting tomorrow, leaving only Flash-Lite. The cut removes Flash and Pro entirely from non-subscribers, where Gemini 3.5 Flash-Lite becomes the only option. Paid plans change too: the $4.99/month tier loses Pro while the $19.99/month tier keeps Flash, Pro, and Deep Think, and users also get a new low/medium/high effort setting tied to usage limits.
Product Hunt 1 week ago 28
Odyssey 3 is being launched as a foundation world model that generates interactive environments from a prompt and predicts their changes as an agent acts. Its Pro version reports a Physics-IQ Verified video-to-video score of 66.1 (best-of-8). Access shifts by offering a research preview and API access for developers, plus adapted uses for robot arms, humanoids, and driving.
MarkTechPost 1 week ago 17
Perplexity released pplx-embed-v2-late, a pair of ColBERT-style multimodal embedding models that retrieve text, images, and rendered PDF pages from a shared embedding space. The 9B model scored 92.4% on MADQA, while a 0.6B query model can search a 9B index to recover about half the quality gap at the 0.6B query cost. The update changes deployment by making both models available on Hugging Face under the MIT license for self-hosting, with a hosted API planned but not yet live.
SiliconANGLE 1 week ago 14 ● 10 sources
OpenAI said its GPT-6 models will shift ChatGPT away from text-only answers by adding an “Intelligent UI” that can insert visuals like charts, forms, buttons, and graphics directly into responses. The feature starts rolling out today for ChatGPT Plus, Pro, Business, and Enterprise subscribers, with free users and Go subscribers getting it tomorrow. ChatGPT will use the new interface selectively on more complex requests, while users can still ask for an interactive response even for prompts that would otherwise get plain text.
The daily briefing
TLDRocket reads all relevant sources, removes duplicate coverage, and summarises the day in two minutes. Follow companies and topics for alerts, or get the briefing in Slack. Free, no spam, unsubscribe anytime.
TLDRocket
17 October 2026
AI governance moved from slideware to the boring-but-important mechanics of who gets to fetch, act, and prove. Meta quietly rewrote parts of its commercial terms, removing wording that appeared to let it use web content to “train” and “maintain the integrity” of its products; the update followed legal scrutiny that it could be read as broad scraping permission. At the same time, users reported that major websites are blocking AI agents from completing tasks across services—turning “agentic” demos into something closer to permissioned automation. ServiceNow is now trying to sell the glue for that reality with its AI Control Tower, positioning it as an enterprise platform to discover and govern agents across models and clouds, while a field study of 15 large-enterprise customers questioned how fully its runtime enforcement and AI lifecycle evidence hold up versus alternatives.
Read the full briefing →