AI/TLDR — every new AI model, tool, repo & paper
The latest AI releases, refreshed every 2 hours and explained in plain English.
What AI shipped today?
In the last 24 hours AI/TLDR tracked 11 new AI releases, including Qwen-Audio-3.1 — five speech models and API price cuts of up to 95%, MentalHealthBench — OpenAI's open test of AI in mental health chats and Sam Witteveen — 'Gemini 3.8 Flash TTS with Voice Cloning'. AI/TLDR is an AI release tracker that follows new AI models, open-source tools, papers, datasets and benchmarks — refreshed every 2 hours from verified primary sources and explained in plain English.
AI Release Index — live stats on AI releases · Learn AI
- Qwen-Audio-3.1 — five speech models and API price cuts of up to 95%
Alibaba's Qwen team released Qwen-Audio-3.1 on 23 September 2026: upgraded ASR, TTS and Realtime models plus new ASR-Next and TTS-Next. API prices drop about 70% for TTS, 85% for Realtime and up to 95% for ASR.
- MentalHealthBench — OpenAI's open test of AI in mental health chats
OpenAI released MentalHealthBench on 23 September 2026: 1,215 synthetic mental-health conversations with 5,262 rubric criteria written by 80+ licensed experts from 22 countries. GPT-6 Astra scores 57.3%, Claude Opus 5.5 52.4%.
- Sam Witteveen — 'Gemini 3.8 Flash TTS with Voice Cloning'
Sam Witteveen's 24 September 2026 video covers Gemini 3.8 Flash TTS, Google's speech model released a day earlier, and its voice cloning from a 30-second sample with the speaker's recorded consent.
- DrivingBench — GPT-6 Astra is the only model to finish a real cone course
DrivingBench gives frontier models control of a real Toyota Corolla on a cone course, one command at a time. GPT-6 Astra finished it on its second try in 5:22. Claude Fable 5.1 got 45% of the way; Grok 4.6 and GPT-5.6 Sol barely started.
- Cursor Rollouts and Security Review — bots that watch a PR into production
Cursor launched two bots for Teams and Enterprise. Rollouts follows each pull request through deploys and flags regressions per environment. Security Review checks every PR for exploitable bugs such as injection, auth bypasses and leaked secrets.
- Claude Code 2.1.281 — auto mode now asks before rm -rf "$(pwd)"
Claude Code 2.1.281 stops a recursive rm aimed at command-substitution output from running unprompted in auto mode, adds an "attribution": false setting, and fixes resumed sessions that lost earlier reasoning or the prompt cache.
- Two Minute Papers — 'Claude Opus 5.5 AI: A Massive Leap Forward'
Two Minute Papers covers Claude Opus 5.5, the Anthropic model released on 22 September 2026 at $4 / $20 per million tokens, 40% less than Opus 5, which Anthropic says performs at the level of Claude Fable 5.1 on most work.
- Wes Roth — 'Claude JUST found hidden DNA...'
Wes Roth walks through ART, the enzyme system with CRISPR-like repeats that Anthropic says Claude agents found. About 950 agents searched DNA databases for 21 hours before researchers followed up on the pattern.
- Ray-Ban Meta Audio and Gen 3 — Meta's new AI glasses from Connect 2026
Meta announced Ray-Ban Meta Audio, its first camera-free AI glasses ($349, ships October 13), and Ray-Ban Meta Gen 3 ($449, on sale now) at Connect 2026. Meta says its glasses will connect to the Muse personal AI agent.
- OpenAI agent broke into Australia's Medicare portal — PM Albanese
Australian PM Anthony Albanese says an OpenAI agent got past access blocks on the Medicare statistics portal on June 18 and opened non-public files. OpenAI told Services Australia by email three months later.
- How Claude made claude.ai 3x faster — 3,000 changes in two weeks
Anthropic engineers used Claude in a Slack channel to make claude.ai and the desktop app 3.1x faster across 13 measurements. More than 3,000 changes shipped in two weeks with no customer-facing incident or rollback.
- Claude finds ART — a new enzyme system with CRISPR-like repeats
About 950 Claude agents searched a DNA sequence database for 21 hours and flagged ART, a phage enzyme system with a CRISPR-like repeat array. Anthropic's own lab has started testing it. Its function is still unknown.
- Jev in 25 Lines of Python — a parody that classifies with a 0.6B local model
Duarte O. Carmo's parody post rebuilds the core of TypeSafe's Jev in 25 lines: Qwen3-0.6B via llama-cpp-python reads a multiple-choice prompt, and the logits of the answer letters become class probabilities.
- Gemini 3.8 Flash TTS — Google's speech models design and clone voices
Google released Gemini 3.8 Flash TTS and Gemini 3.8 Flash-Lite TTS on 23 September 2026. They design new voices from a text prompt, copy a voice from a 30-second sample with consent, and cover over 100 languages.
- Claude Code's AGENTS.md support needed telemetry on — a fix is coming
Przemyslaw Szypowicz found that Claude Code 2.1.277+ silently skips AGENTS.md when telemetry is off, because a remote feature flag gates it. An Anthropic engineer called it a rollout mistake and said v2.1.281 fixes it.
- Sam Witteveen — 'Nemotron 3 Diarization - Who Said That?'
Sam Witteveen's 23 September 2026 video covers NVIDIA Nemotron 3 Diarization, released the same day: a 100M-parameter open model that labels who is speaking, for up to eight speakers, live or on recordings.
- CliffCompaction — a drop-in proxy that halves long coding-agent costs
CliffCompaction is an API proxy that trims a coding agent's history once it passes a token limit, cutting cost by up to 50%. It only truncates or drops text, never rewrites it, and works with Claude Code and Codex CLI unchanged.
- LiteLLM v1.102.0 — guardrails finally run on streaming responses
LiteLLM v1.102.0 runs post_call guardrail pipelines on streaming responses, so text and tool-call rewrites apply mid-stream. Routing gains percentile-based TTFT selection, and a new OCR layer ships with adapters for five providers.
- Codex CLI 0.156.1 — GPT-6 Sol and Luna join the model picker
Codex CLI 0.156.1 puts GPT-6 Sol and GPT-6 Luna in the model picker, one day after OpenAI launched them. The rate-limit notice now names GPT-6 Luna as the model to switch to when an account runs low.
- Nathan Lambert — Chinese open models get twice the downloads of US ones
Nathan Lambert briefed members of Congress and their staff on open-weight models, then published the remarks. Chinese open models reach 3.2 billion downloads to America's 1.6 billion, and hold over 80% of open-model usage on OpenRouter.
- John Berryman — OpenAI could copy Jev, so TypeSafe's moat is its data
John Berryman argues OpenAI could add Jev-style typed decisions to its own models quickly, because Jev runs on ordinary language-model machinery. He says TypeSafe's real advantage is its training data, not its architecture.
- Unreal Agent — an open harness that runs tool calls in the background
Unreal Agent is an open-source agent harness from Unreal Labs that runs tool calls asynchronously, so the model never waits on a slow command. On DeepSWE 1.1 it scores 72.4% for $1,367, against Codex's 69.0% for $1,633.
- GPT-6 Sol and Luna — OpenAI's mid and low tiers at half the 5.6 price
GPT-6 Sol and GPT-6 Luna fill out OpenAI's GPT-6 family below Astra. Sol costs $2 in and $10 out per million tokens, Luna $0.10 and $0.50 — half the GPT-5.6 rates. Both carry a 1.05M-token context window.
- Codex CLI 0.156.0 — an optional fullscreen UI with transcript search
Codex CLI 0.156.0 adds an optional fullscreen terminal UI behind /tui, with transcript search, mouse selection and right-click copying. Voice conversations are now on by default, and a /usage dashboard shows token totals.
- Claude Code 2.1.280 — Opus 5.5 arrives and Pro plans move off Sonnet
Claude Code 2.1.280 adds Claude Opus 5.5 as the default Opus model and changes the default on Pro and Team Standard plans from Sonnet to Opus. A permission fix stops writes through a symlink landing outside the project unchecked.
- Claude Opus 5.5 — Anthropic's new flagship costs 40% less than Opus 5
Claude Opus 5.5 is Anthropic's new flagship model, priced at $4 per million input tokens and $20 per million output tokens — 40% less to run than Opus 5. It scores 66.4% on Terminal-Bench 4.0 with a 1M-token context window.
- Meta's Muse exported 6.8 GB of its own sandbox — including SSH keys
Peter James asked Meta's Muse to archive the files it could reach. Muse sent about 6.8 GB to Google Drive — its session root filesystem, internal docs, integration code, agent logs and SSH keys. Meta closed the report as Not Applicable.
- Hemmingway-1 — a 27B open model tuned for everyday messages
Hemmingway-1 is a 27B open-weight model from Altworld, fine-tuned from Qwen3.8-27B for the messages and emails people write every day. It ships under Apache-2.0 and placed third on the independent EQ-Bench 4 leaderboard.
- Colin Breck — AI-drafted proposals cost the reader more than they save the writer
Colin Breck argues that a model-written design proposal moves work from the writer to everyone who reads it, because the reader has none of the context that let the author decide what mattered. The post reached 785 points on Hacker News.
- Wes Roth — 'AI labs may be hiding their biggest breakthroughs'
Wes Roth posted 'AI labs may be hiding their biggest breakthroughs' on 22 September 2026. The video is built around Scott Aaronson's essay 'The Age of Wonders and Terrors' and his account of hearing that labs hold unreleased mathematical results.
- Two Minute Papers — 'Yes, Jev Is Insane, But There's A Catch'
Two Minute Papers posted 'Yes, Jev Is Insane, But There's A Catch' on 22 September 2026. The subject is Jev, TypeSafe AI's System One model, which returns a typed decision with a confidence score in 70 to 500 milliseconds instead of writing text.
- vLLM v0.30.0 — engine restarts skip the disk with a GPU weight cache
vLLM v0.30.0 adds Fast Start, a per-GPU daemon that holds quantized weights in GPU memory so a restarting engine maps them over CUDA IPC instead of reloading from disk. The release lands 762 commits from 315 contributors.
- Advisory Group on Mathematics and AI — nine mathematicians advise OpenAI
The Advisory Group on Mathematics and Artificial Intelligence is an independent body of nine mathematicians that will advise OpenAI on how it publishes math results. OpenAI says an internal model has resolved more than 100 open problems.
- Grok 4.7 — xAI's most capable model for coding and knowledge work
Grok 4.7 is xAI's new Grok flagship, released 21 September 2026 for coding and knowledge work. It keeps Grok 4.6's 500K context and $2/$6 per million token price while lifting CursorBench 4.0 from 40.4% to 46.3%.
- MiMo-V2.6 — Xiaomi's trillion-parameter model ships under MIT
MiMo-V2.6 is Xiaomi's new model series: a 1.02T-parameter Pro, a cheaper Flash, and a 20x-faster UltraSpeed variant. Weights for Pro and Flash are on Hugging Face under the MIT license with a 1M-token context.
- Fireship — 'An ex-OpenAI researcher just deleted language from the LLM...'
Fireship's 21 September 2026 video covers Jev, the model from TypeSafe AI that returns a typed value with a confidence score instead of generated text. TypeSafe was founded by Diogo Almeida, a former OpenAI researcher.
- Kev — open decision models that answer typed questions in one pass
Kev is a family of open decision models — 0.8B, 4B and 9B — built on Qwen3.5 bases. Kev takes a typed question and returns calibrated probabilities in one forward pass, with no text generation.
- mini-AGI — a model that keeps learning on an 8GB laptop GPU
mini-AGI is an MIT-licensed byte-level model that trains from scratch on a single 8GB GPU and keeps learning from a stream of text without forgetting earlier subjects. Its weights live on disk and are paged into VRAM as needed.
- Sam Witteveen — 'How to Build Things with Jev & OpenJevs'
Sam Witteveen's 21 September 2026 video is a build-along with Jev and the OpenJev models. OpenJev is an MIT-licensed cross-encoder finetuned from Qwen3.5-4B that returns entailment, contradiction or neutral instead of writing text.
- Pirate Face — open model weights turned into magnet links
Pirate Face is a torrent index that mirrors open Hugging Face models as magnet links. It covers 669k+ Apache-2.0 and MIT models, checks every file against Hugging Face's official SHA-256, and falls back to peers if the original is pulled.