Skip to content
DigIntoAI/Model intelligence
Model intelligence feed

Model intelligence.

Follow the changes. Understand what comes next.

Editor’s pickOther providers
Other providers/Steam

Valve's Answer to AI Slop and Clone Games on Steam: We Won't Police Developers' Tools

At Steam Indie Connect in Korea, Valve staff addressed AI slop and clone games head-on: no gatekeeping on AI tools, and the best defense against clones is community and steady updates.

Latest stories53

Newest published
25

Atria Dawn Formal Weights Are Out: 512K Context on Hugging Face

Shanghai Artificial Intelligence Laboratory has published formal open weights for Atria Dawn on Hugging Face (internlm/Atria-Dawn), with an FP8 sibling. The formal release lists a 512K context window, versus 256K for the earlier Preview. The model card describes a research- and engineering-oriented agent built on the GLM-5.2 MoE foundation—it is not a new Zhipu GLM generation under another name. Weights are text-only and very large on disk; that file size is not a minimum VRAM figure.

Atria Dawn / Atria-Dawn-FP8Read story
26

Replit Monitoring Agent Can Investigate, Fix and Auto-Publish Production Issues

Replit introduced Monitoring Agent—an autonomous on-call for published apps. It can read production logs and metrics, investigate an alert, make and validate a fix, and email a summary. With auto-publish on, it can deploy a validated fix to production. Earlier App Monitoring covered health checks and downtime alerts; this release adds the fix-and-ship step. Enable monitoring alerts and auto-fix in Publishing settings; auto-publish is optional. Separate from Replit Agent’s default model choice.

Monitoring AgentRead story
27

Block Releases Dessa, a Full-Duplex Speech Model Trained Only on Speech

Block’s small AI research lab shipped Dessa: speech in, speech out, full-duplex. It says the model was trained from scratch on human conversation audio only—no text, no pretrained base. Research preview for now: a small group of AI safety researchers get early access; everyone else joins a waitlist on the official page. Block says inference beats real time and the model runs locally on a MacBook Pro M4 Max or better. Training, per the page, used one NVIDIA GB200 NVL72.

DessaRead story
28

Simular Agent S4 Evolves Interfaces Without Retraining the Model

On October 21, 2026, Simular introduced Agent S4, which evolves the interface repository computer-use agents work through while the task agent’s model and harness stay fixed. With Qwen3.8-27B and an evolved repository (no scratch VM), it reaches 21.3% binary success on the full 108-task OSWorld V2 set—about 3.8× the same model with fixed interfaces (5.56%), and above GPT-5.5 (13.0%) and Claude Opus 4.8 (20.6%) at a small fraction of their reported task cost. Evolving an interface repository is not the same as Cursor’s self-hosted desktop computer use.

Agent S4Read story
29

UK Opens World's First AI Biosecurity Lab

On October 24, 2026, the UK Cabinet Office announced the AI Biosecurity Lab (AIBL)—the world’s first dedicated AI biosecurity organisation, per the government. It has £100 million in government funding and up to £200 million in commitments from OpenAI and Anthropic (combined package up to about £300 million). Compute and model credits go to the lab, not to personal API or coding quotas. Partners also include Google DeepMind and Basecamp Research. AIBL will be an independent non-profit focused on earlier detection of biological threats and faster medical countermeasures.

AI Biosecurity Lab (AIBL)Read story
30

Brotli Unauthenticated RCE Affects Node, Bun and Envoy

CVE-2026-111497 is an out-of-bounds write in the Google Brotli decoder. Crafted compressed input can crash a process, leak memory, or enable remote code execution when an app decompresses attacker-controlled Brotli data. Fixed lines include Node.js 22.23.4 / 24.22.1 / 26.12.1, Bun 1.4.4, Envoy 1.40.0 and backports, Istio patched lines, and upstream Brotli 1.0.10 / 1.1.1 / 1.2.1. OpenAI Codex Security researchers reported the issue; Patch the Planet tracks mitigations. Until you can upgrade, disable Brotli decompression of untrusted data.

Brotli / CVE-2026-111497Read story
31

Sakana AI Announces Aquarium, an Ontology-Based Agent Platform

On October 23, 2026, Sakana AI announced Sakana Aquarium, a platform to build, run, and operate organization-specific AI agents on a shared ontology. It groups Ontology (structured knowledge, rules, and data), Governance (permissions, pre-execution approval, audit trails), and Managed Agents / Agent Runtime (tasks that reference the ontology). Workflows—including human approval steps—can be described in YAML. Public pricing, a self-serve open date, and general availability by region are not listed; the product page points to a contact form. We have not tested Aquarium yet.

Sakana AquariumRead story
32

Sage-A1.5 Open Weights: Self-Verification Is Opt-In and Off by Default

InternScience’s Sage-A1.5 is a ~35B Mixture-of-Experts agentic model with Apache 2.0 weights on Hugging Face (also ModelScope and an FP8 build). Self-verification and multi-agent orchestration are off by default—the former needs an explicit chat-template flag; the latter needs your harness to run subagents. Native context is 256K (512K via YaRN). The arXiv report appeared around Beijing time October 21, 2026, early morning. Scores on the model card are the authors’ claims; we have not tested the model yet.

Sage-A1.5Read story
33

SCNet Offers ¥0.10 Token Packs for DeepSeek and GLM to New Users

China’s National Supercomputing Internet (SCNet) MaaS page is running a new-user flash sale: DeepSeek-V4.1-Flash-1017 and GLM-5.3-Flash token packs at ¥0.10 each, 50 million tokens per pack. The campaign runs through Beijing time 2026-11-10 23:59. This is a domestic supercomputing-channel promo, not a global DeepSeek or Zhipu announcement.

SCNet MaaS Token Packs / DeepSeek-V4.1-Flash-1017 / GLM-5.3-Flash / Token PlanRead story
34

Baseten Mammoth-1 and Elephant-1 Are Free on Vercel Through October 27

Baseten’s Base Labs released mammoth-1 (27B) and elephant-1 (9B) decision models—Choice, Score, and Boolean questions that return probability distributions, not generative coding LLMs. They are free on Vercel AI Gateway through October 27, 2026 (also free for a week on Baseten Model APIs), open-source on Hugging Face as baselabs/mammoth-1 and baselabs/elephant-1, and post-trained from Qwen3.5-27B / Qwen3.5-9B. Rankings are Baseten’s claims citing Jev and Artificial Analysis.

Mammoth-1 / Elephant-1 / System One / Vercel AI GatewayRead story
35

Canva's First Image Model Craft Hits Dream Lab and a Public API

Canva Research’s first image model, Craft, is available to Canva users globally through Dream Lab and to developers via a public image-generation API with model craft_v1_2k. DesignArena lab and model rankings are Canva’s own citations. Canva MCP for Craft is listed as Coming Soon.

Craft / Dream Lab / craft_v1_2kRead story
36

Framer MCP Connects Claude, Codex, and Cursor — Publish Still Needs a Go-Ahead

On about October 22–23, 2026 Beijing time, Framer shipped Framer MCP: a remote Model Context Protocol server so agents can work on the canvas, CMS, and project settings without keeping the Framer editor open. Live-site updates still require an explicit publish. We have not tested the flow.

Framer MCPRead story