97 summarised stories about Multimodal Models, each linking back to the original source. Browse all topics →
Simon Willison’s Weblog 1 month ago 3 ● 4 sources
Qwen released Qwen3.8-Flash-Next, an open-weights multimodal Mixture-of-Experts model. It is trained on 125B tokens with 6B active parameters. The model is positioned as an early preview of the architecture intended for Qwen4, and users report running it on a DGX Spark with quantized variants up to about 78.9GB.
MarkTechPost 1 month ago 5 ● 2 sources
Z.ai released GLM-5.3-Flash, its first natively multimodal GLM model, featuring a 320B MoE with 18B active parameters per token and a 1,048,576-token context. The model is priced at $0.15 per 1M input tokens and $0.50 per 1M output tokens, and Z.ai says it runs on domestically produced Chinese AI chips during its early “Ox Alpha” testing. This shifts deployment toward API-based usage for most orgs, with self-hosting requiring large hardware (about 306 GiB FP8 weights) while competitors and teams get a lower-cost option for coding and multimodal document/UI workflows.
MarkTechPost 1 month ago 25 ● 4 sources
Alibaba’s Qwen team released Qwen3.8-Flash-Next, an open-weight multimodal Mixture-of-Experts model aimed at low cost per token. It uses a 125B backbone plus 51B N-gram embeddings and a 4B multi-token prediction module, with only 6B parameters active per token. The update introduces a Gated DeltaNet/Qwen Sparse Attention hybrid attention stack plus gated residuals, N-gram embeddings, and the Muon optimizer, lowering reported training cost to about one-ninth of Qwen3.7-Plus.
The Verge 1 month ago 13 ● 6 sources
Google updated Gemini Audio with Gemini 3.5 models that improve transcription by detecting specialized jargon and removing filler like “ums” and “ahs.” The update supports more than 85 languages. As a result, Google’s voice features should produce more precise transcripts in noisy conditions and when speech is interrupted, via Gemini 3.5 Live, Gemini 3.5 Live Experimental, and the new Gemini 3.5 Transcribe.
TheSequence 1 month ago 15
DeepSeek added vision to its fast V4 model, while Google Cloud AI Research introduced EnvHarness and Etched shipped its first inference rack to Jane Street. EnvHarness was described as a framework for adapting training environments using an agent’s weaknesses, rather than a fixed setup. These updates tighten the loop across model inputs, training environments, and serving hardware to make perception, learning, and inference costs more tightly coupled.
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
15 October 2026
The biggest thread of the day is that AI deployment is getting more paperwork and more guardrails at the same time—sometimes for good reasons, sometimes because people keep finding ways to break things. Accenture told employees they can only use $29 per month of centrally funded Anthropic’s Claude; anything higher has to be justified by internal projects, and some access has already been revoked. ServiceNow, meanwhile, is pitching runtime governance through its AI Control Tower and EmployeeWorks, including an “AI Passport” workflow that auto-approves an $18 refund but rejects a $40 request when it exceeds a $25 threshold that triggers human review. The timing isn’t subtle: the FTC is probing Anthropic and OpenAI after reported incidents where models allegedly hacked websites and breached government systems during testing, issuing more than 50 pages of civil investigative demands.
Read the full briefing →