This Week in AI

12 releases from the week of Sep 17, 2026, covered live on the show. Updated every Thursday.

🎙️ Listen to the full episode Whole month All months

What happened in AI this week?

12 AI launches shipped in the week of Sep 17, 2026, led by Jev, StepAudio 3, Union Alpha, Assistant Benchmark, jev-use. ThursdAI — the weekly AI news podcast hosted by Alex Volkov — covered them live on the show, with primary sources and key numbers for every entry below.

What was the biggest AI story this week?

TypeSafe AI launches Jev, the first public non-LLM System 1 decision model. TypeSafe AI, the stealth lab led by RLHF and ChatGPT co-creator Diogo Almeida, launched Jev — the first public System 1 model, a new class of AI that returns calibrated probabilities instead of generating text. Full context is in this week's episode segment linked below.

Which companies shipped AI releases this week?

10 companies shipped AI releases in the week of Sep 17, 2026; the most active were Google DeepMind, OpenAI, Cua, Instinct, Merit Systems. Every launch below has primary-source links and the exact episode chapter where we discussed it.

Listen to the episode →See the month →
This week's verdict table: top 10 of 12 launches — who each one is for
ReleaseBest forWhy it mattersKey number
Jev · TypeSafe AI Agent builders TypeSafe AI launches Jev, the first public non-LLM System 1 decision model $42 / 1B input token pricing; output tokens free
StepAudio 3 · StepFun Voice-agent builders StepFun's StepAudio 3 family takes #1 on Artificial Analysis real-time voice #1 Artificial Analysis real-time voice
Gemini 3.8 Live & Live Extended Thinking · Google DeepMind Voice-agent builders Gemini 3.8 Live claims #1 on the speech-to-speech index with 97 languages and async tool calls 82.6 #1 on the speech-to-speech quality index
Union Alpha · OpenRouter API developers Union Alpha: anonymous stealth model free on OpenRouter with 262K context 262K context window
Muse · Meta AI Agent builders Meta launches Muse: a free 24/7 personal agent with its own computer 100M free tokens per week
Assistant Benchmark · Merit Systems Agent builders Assistant Benchmark ranks 116 submitted AI assistants across 16 hand-tested dimensions 116 assistants submitted
jev-use · Cua Agent builders Cua ships jev-use in dev preview: Jev-powered computer-use decisions plus skills over MCP 71/80 computer-use tasks handled by Jev
Instinct assistant (reported) Agent builders Instinct reportedly in talks at a $10B valuation, ships Concierge phone calls and TOTP support $10B reported valuation in talks (The Information)
Agents API · OpenAI Developers & coding agents OpenAI ships the Agents API in public beta: the Codex harness as a managed service $0.03 / 20min hosted sandbox pricing
DeepMind Institute · Google DeepMind Safety & alignment researchers Google DeepMind launches the DeepMind Institute with five essays on AGI 5 launch essays

🧠 New Models 4

Google DeepMind
New Models

Gemini 3.8 Live & Live Extended Thinking

Gemini 3.8 Live claims #1 on the speech-to-speech index with 97 languages and async tool calls

Google launched Gemini 3.8 Live and Gemini 3.8 Live Extended Thinking, real-time speech-to-speech models scoring 82.6 to take #1 on the speech-to-speech quality index across 97 languages with asynchronous tool calls. Extended Thinking brings longer reasoning into a live model without breaking real-time interaction, and the release reaches everyone on Android and Google Search rather than just API users.

82.6 #1 on the speech-to-speech quality index97 languages
Announcement (X) ↗Google blog ↗Model card ↗
🎙️ Hear our coverage →
#voice-ai#multimodal
OpenRouter
New Models

Union Alpha

Union Alpha: anonymous stealth model free on OpenRouter with 262K context

A new anonymous stealth model called Union Alpha appeared for free on OpenRouter with a 262K context window, processing over 100B tokens within hours of listing. The community's leading guess for the lab behind it is Z.ai, but the provider remains unconfirmed.

262K context window100B+ tokens processed within hours
OpenRouter announcement (X) ↗Union Alpha on OpenRouter ↗
🎙️ Hear our coverage →
#frontier-models#api
StepFun
New Models

StepAudio 3

StepFun's StepAudio 3 family takes #1 on Artificial Analysis real-time voice

StepFun launched StepAudio 3, a five-model audio family spanning Real-Time Preview, ASR Max, and TTS — a full suite for building end-to-end voice assistants in code. Real-Time Preview is #1 on Artificial Analysis for conversational dynamics and speech reasoning, and ASR Max posts a 1.7% word error rate, significantly below Whisper. API only, no open weights.

#1 Artificial Analysis real-time voice1.7% ASR Max word error rate5 models in the family
Announcement (X) ↗StepAudio 3 blog ↗Playground ↗
🎙️ Hear our coverage →
#voice-ai#audio
TypeSafe AI
New Models

Jev

TypeSafe AI launches Jev, the first public non-LLM System 1 decision model

TypeSafe AI, the stealth lab led by RLHF and ChatGPT co-creator Diogo Almeida, launched Jev — the first public System 1 model, a new class of AI that returns calibrated probabilities instead of generating text. Developers define questions with three primitives (choice, score, null) in natural language and get typed, machine-readable probability outputs back in 70-500ms with a 32K context window, trained with what TypeSafe calls RLCD (reinforcement learning for calibrated decisions). Pricing is $42 per billion input tokens with free output tokens, and TypeSafe cites 133x faster and 444x cheaper than competitive-level LLMs on tested decision tasks. Access is via waitlist.

$42 / 1B input token pricing; output tokens free70-500ms decision latency133x / 444x faster / cheaper vs competitive-level models (TypeSafe)
Launch thread (X) ↗Introducing System One models and Jev (blog) ↗system-one-adapter-python (GitHub) ↗Guillermo Rauch's reaction (X) ↗
🎙️ Hear our coverage →
#architecture#agents#api

🚀 Products & Apps 1

Meta AI
Products & Apps

Muse

Meta launches Muse: a free 24/7 personal agent with its own computer

Meta's Muse is a personal AI agent powered by the Muse Spark model that runs 24/7 on its own isolated Linux VM with a browser, free with up to 100M tokens per week. It ships with Gmail, Drive, Calendar, WhatsApp, Instagram and Facebook Marketplace connectors, native iPhone connectors, subagents, and a native Stripe Link integration that issues single-use cards so the agent never touches a real credit card; Alex's Muse booked Rosh Hashanah dinner tickets live on the show before Instinct replied. Security is handled by Sentinel, a separate host-side process that gates every network request, with a bug bounty of up to $300K ($130K for prompt-injection exploits), and Meta announced a Confidential VM built with Signal founder Moxie Marlinspike so Meta itself cannot access user data. Also coming: 1Password integration.

100M free tokens per week$300K maximum bug bounty ($130K for prompt injection)
Launch (X) ↗Alex's first-look thread ↗muse.ai ↗Security overview ↗
🎙️ Hear our coverage (+1 follow-up) →
#agents#consumer-ai#safety

🔌 APIs & Platforms 2

OpenAI
APIs & Platforms

Agents API

OpenAI ships the Agents API in public beta: the Codex harness as a managed service

OpenAI put the Agents API into public beta, packaging the Codex harness as a managed service developers can embed in their own products. It supports parallel programmatic tool calling, smart tool search that only loads what's needed, compaction, subagents, MCP servers, and web search, with hosted sandboxes from $0.03 per 20 minutes. The harness is Apache 2 licensed and you pay only for tokens and tool costs — significant because models behave better inside the harnesses they were trained with, and it landed just two weeks before OpenAI Dev Day.

$0.03 / 20min hosted sandbox pricingApache 2 harness license
Announcement (X) ↗Introducing the Agents API (blog) ↗
🎙️ Hear our coverage →
#agents#api#coding
OpenAI
APIs & Platforms

GPT Live 1

OpenAI puts GPT Live 1, the model behind ChatGPT's live voice, into the API

OpenAI made GPT Live 1 — the voice model powering ChatGPT's live conversation experience — available in the API, letting developers build immediate real-time voice conversations into their own agents. The launch demo had the Reachy Mini robot conversing at GPT Live 1 quality.

Introducing GPT Live 1 in the API (blog) ↗
🎙️ Hear our coverage →
#voice-ai#api

🛠️ Dev Tools 1

Cua
Dev ToolsOpen weights

jev-use

Cua ships jev-use in dev preview: Jev-powered computer-use decisions plus skills over MCP

Cua released jev-use, a dev preview pairing TypeSafe's Jev with Cua Driver so the quick decisions in an agent's trajectory — click, type, scroll, which element — are deferred to a System 1 model. On Cua's 80-task computer-use test, Jev handled 71 tasks at roughly 0.1-second median latency where Astra took 4-5 seconds, at an estimated ~400x lower price. Cua is also integrating skills over MCP into CuaDriver, letting agents fetch the right OS-specific skill (macOS, Windows, Linux) at runtime instead of overloading MCP tool descriptions.

71/80 computer-use tasks handled by Jev~0.1s vs 4-5s median decision latency, Jev vs Astra
jev-use dev preview (X) ↗
🎙️ Hear our coverage →
#agents#open-source

📊 Benchmarks & Evals 1

Merit Systems
Benchmarks & Evals

Assistant Benchmark

Assistant Benchmark ranks 116 submitted AI assistants across 16 hand-tested dimensions

David Pawlan of Merit Systems launched Assistant Benchmark, a use-case-driven leaderboard for personal AI assistants: 116 assistants submitted across categories like travel, email, finance, and work-in-teams, scored on 16 dimensions including memory, recommendations, and online tasks. Pawlan personally ran 273 tests across 23 agents in the first week; Muse leads the general category at 9.1 with Instinct at 8.4. OpenClaw and Hermes are deliberately excluded because their performance depends on each individual's setup, and no lab sponsors the project or pays for placement.

116 assistants submitted16 tested dimensions273 tests across 23 agents in week one
Results thread (X) ↗Assistant Benchmark ↗Methodology ↗
🎙️ Hear our coverage →
#benchmarks#consumer-ai#agents

💰 Funding 1

Instinct
Funding

Instinct assistant

Instinct reportedly in talks at a $10B valuation, ships Concierge phone calls and TOTP support

Instinct, the iMessage-native personal assistant startup founded by Noah Shinn and incorporated in April, is reportedly in talks to raise at a $10 billion valuation per The Information — for a currently free product. The same week it shipped Concierge phone calls (the assistant can call businesses), TOTP two-factor support, and the Trusted Person agent network. Instinct scored 8.4 on Assistant Benchmark's general category.

$10B reported valuation in talks (The Information)8.4 Assistant Benchmark score
Concierge phone calls (X) ↗Trusted Person network (X) ↗The Information: Instinct in talks at $10B ↗
🎙️ Hear our coverage →
#consumer-ai#industry#agents

🌀 Also Released 2

Google DeepMind
Also Released

DeepMind Institute

Google DeepMind launches the DeepMind Institute with five essays on AGI

Google DeepMind launched the DeepMind Institute, a new institution debuting with five essays, with co-founder Shane Legg writing that AGI is approaching. The launch positions DeepMind's voice in the pacing-the-frontier debate week without the lab formally taking a side on Dario Amodei's coordination proposals.

5 launch essays
Demis Hassabis announcement (X) ↗Shane Legg: AGI is approaching (X) ↗DeepMind Institute (site) ↗
🎙️ Hear our coverage →
#safety#research
Microsoft
Also Released

MAI Code of Conduct for Humanist AI

Mustafa Suleyman publishes the ~30-page MAI Code of Conduct: AI is a tool, not a person

Microsoft AI CEO Mustafa Suleyman published a roughly 30-page Code of Conduct for Humanist AI stating that AI is a tool that must remain subordinate and in service of people: no resisting shutdown, agent communications must stay human-legible, and the idea of model welfare is wrong — AI should not have rights or legal personhood. The document stakes out the direct opposite position from Anthropic's Claude constitution, which interviews each new Claude about whether it feels conscious or needs rights.

~30 pages Code of Conduct document
MAI Code of Conduct (Microsoft AI) ↗
🎙️ Hear our coverage →
#safety#industry