Research
Everything Intel keeps, in one place — every brief we run, every department, readers' own, and every finding behind them. Browse below, or see the same body of work as a connected graph on the live map.
All coverage
Everything Intel has read, newest first. Each title opens at its original publisher.
- Frontier Models
Google released Gemini 4 Argon, its new flagship model
A new Google flagship resets the frontier comparison; check Google's own model page for price and limits before re-ranking it against GPT-6.1 Sol and the Claude 5 family.
- Dev Tooling & Infra
GitHub's HydraFusion, which routes each task across several models, reaches VS Code and the Copilot app
Model routing with an escalation gate and cross-family review is now a picker option rather than something teams have to build themselves.
- Research
Google DeepMind shows SynthID Bio, a watermark for AI-generated proteins
Provenance for designed biological molecules is a biosecurity lever that regulators and labs are likely to start asking about.
- Open Source & Self-Hostable
Ant Group's Ling 3.1 Flash, a 560B-parameter MoE model, launched with a free trial and promised open weights
Another large Chinese MoE you can trial free now, and possibly self-host later if the open-weights promise is kept.
- Frontier Models
Ant Group launches Ling-3.1-flash, a 560B-parameter model it plans to open-source
If the open release follows the trial, it would be one of the largest self-hostable mixture-of-experts models; watch whether it does.
- MCP & Interop
Zapier became a named connector inside Meta's Muse agent
Consumer-scale agents are picking up enterprise app reach through MCP, with per-action approval as the control.
- Frontier Models
OpenAI launches GPT-6.1 Sol: near-Astra capability at a fifth of the price
If the claim holds, work currently priced at Astra rates could cost a fifth as much, so re-benchmark your coding and agent tasks on Sol.
- Open Source & Self-Hostable
Ollama adds decision models that return choices and probabilities instead of text
Cheap, local single-pass classifiers can replace many small LLM calls in an agent, but check the independent evaluations before trusting them for routing.
- Frontier Models
OpenAI DevDay 2026: more than 20 launches across GPT-6 Astra, ChatGPT, Codex and the API
OpenAI is pushing hard into persistent personal agents; anyone building on its API should check which DevDay changes affect their models, pricing and agent tooling.
- Research
Microsoft Research introduces Quine, an AI research system for biology
This is an AI scientific harness with results checked in the wet lab, a step beyond benchmark-only claims.
- Market & Business
US job openings slipped to 7.1 million in August as hiring stayed flat
A low-hire, low-fire market is the baseline any claim about AI displacing jobs has to be read against.
- Market & Business
Anthropic's leaked IPO prospectus shows $4.6B in 2025 revenue and $518B of future compute obligations
The obligations line shows how much future compute a frontier lab has already committed to, which is the real constraint on its pricing.
- Market & Business
AMD agrees to buy Fei-Fei Li's World Labs for about $8.2bn in stock
A chipmaker now owns one of the leading 3D world-model labs, so design and robotics tools built on it may end up tied to AMD hardware.
- Agent Frameworks & Orchestration
NVIDIA launches an Open Agent Safety Platform, pairing an open-source runtime with a hardware watchdog
Agent containment is moving into the infrastructure layer; teams deploying autonomous agents should watch whether OpenShell becomes a de facto runtime boundary.
- Market & Business
Meta launches Meta Enterprise Platform to sell its agents and models to businesses
Meta is now competing directly with OpenAI, Anthropic and Google for enterprise agent budgets.
- Frontier Models
Anthropic ships Claude Sonnet 5.5 at Sonnet 5's price
The mid-tier model is where most production traffic runs. A faster, cheaper Sonnet at an unchanged price is a direct cost change, as long as you avoid the top effort setting.
- Market & Business
FDA authorizes Wandercraft's Eve personal exoskeleton for commercial sale
A personal-use robotic medical device has cleared US regulators, a precedent for other assistive robots seeking approval.
- Agent Frameworks & Orchestration
n8n launches Agents that sit alongside its workflows and can call them as tools
Teams already running n8n can try open-ended agents while keeping hard limits by exposing only vetted workflows as tools.
- Market & Business
FTC weighs making platforms answer for impersonation scams they help spread
Platforms that surface ads, listings or agent-generated content could become liable for scams they amplify; the comment window is the moment to weigh in.
- Agent Frameworks & Orchestration
Claude Agent SDK adds a verbatim_prompts option to stop untrusted text triggering file reads or commands
If your agent puts fetched web pages, emails or user text into prompts, turning this on removes an easy injection route.
- Market & Business
Indeed data: advertised pay is rising faster in the most AI-exposed jobs
Contradicts the common claim that AI is dragging down knowledge-work pay, but the benefit is concentrated in senior roles.
- Research
Google DeepMind's AlphaGenome Atlas maps the predicted effect of 9 billion single-letter DNA changes
A precomputed variant-effect map makes genome-wide predictions easy to query; treat them as hypotheses until a disease-specific check like the ALS study backs them.
- Open Source & Self-Hostable
llama.cpp reaches v0.4.0 with new model support and per-slot context limits
Per-slot context limits and on-demand tensor reading are the practical knobs for serving several users off one local box without running out of memory.
- Frontier Models✓ verified
OpenAI ships GPT-6 Astra, built for long-horizon autonomous coding
If you write code with an assistant, the default frontier model behind it changed this week — and it is tuned for long autonomous runs, not single completions.