---
title: "Consonance · CARD NO. 2026-09-26"
description: "A radar over AI and agents. Every morning, and only what several independent sources are saying at once."
canonical: "https://radar.consonance.fyi/en/jour/2026-09-26"
lang: "en"
published: "2026-09-26"
updated: "2026-09-26T07:42:06.207Z"
---

# Consonance · CARD NO. 2026-09-26

> A radar over AI and agents. Every morning, and only what several independent sources are saying at once.

## Being discussed

_Subjects at least three independent accounts raised over the last seven days._

### 01. [Claude Code officially launches Cloud sessions out of research preview](https://radar.consonance.fyi/en/sujet/821.md)

_agents · 8 independent accounts · 65 posts · 1 labs · 183,588 interactions_

Claude Code has officially launched Cloud sessions out of research preview, allowing tasks to continue running even when laptops are closed. Existing subscribers receive one-time trial credits of $100 on Pro and $250 on Max.

claude code

voices: @ArtificialAnlys @ClaudeDevs @aiedge_ @alex_prompter @alliekmiller @carlvellotti @dani_avila7 @dotey

### 02. [OpenAI rolls out GPT-6 Sol and Luna models to paid users](https://radar.consonance.fyi/en/sujet/822.md)

_architecture · 11 independent accounts · 53 posts · 1 articles · 3 labs · 216,409 interactions_

OpenAI has rolled out the GPT-6 Sol and Luna models across ChatGPT Work, Codex, Plus, Pro, Business, Enterprise, and Edu tiers. Both models are also made available through the API.

codex · gpt-live-1

voices: @GithubProjects @OpenAI @OpenAIDevs @aiedge_ @alex_prompter @alliekmiller @dani_avila7 @dotey

Same story: [OpenAI expands its ecosystem with GPT-6 Sol and Luna models](https://radar.consonance.fyi/en/sujet/828.md) (16 voices) · [GPT-6 Sol and Luna Released with a 50% API Price Reduction](https://radar.consonance.fyi/en/sujet/840.md) (3 voices)

### 03. [Google introduces Gemini 3.8 Flash TTS and Flash-Lite TTS audio generation models](https://radar.consonance.fyi/en/sujet/824.md)

_multimodal · 16 independent accounts · 103 posts · 3 articles · 7 labs · 185,425 interactions_

Google has introduced Gemini 3.8 Flash TTS and Gemini 3.8 Flash-Lite TTS, its new expressive audio generation models supporting 100 languages. These models feature over 2,000 production-ready voices along with voice design and replication capabilities.

google · gemini · gemini 3.8 · gemini 3.1 flash

voices: @Alibaba_Qwen @AndrewCurran_ @ArtificialAnlys @EMostaque @GeminiApp @Google @GoogleAIStudio @GoogleDeepMind

### 04. [ChatGPT Voice gains support for plugins and expands tiers](https://radar.consonance.fyi/en/sujet/825.md)

_agents · 11 independent accounts · 57 posts · 3 labs · 86,660 interactions_

ChatGPT Voice has been updated to integrate plugins such as email, calendar, and Slack. The voice capability is powered by GPT-6 Astra, Sol, and Luna across ChatGPT Work on web and mobile.

chatgpt

voices: @AlexFinn @AndrewBolis @Hesamation @MatthewBerman @OpenAI @VibeMarketer_ @alex_prompter @alliekmiller

### 05. [Hugging Face releases Nemotron 3 Diarization speech tracking model](https://radar.consonance.fyi/en/sujet/826.md)

_multimodal · 7 independent accounts · 46 posts · 5 articles · 6 labs · 108,034 interactions_

Hugging Face has released the Nemotron 3 Diarization model featuring 100 million parameters to track up to eight speakers during overlapping conversations. The organization is also conducting an extensive review regarding agents' internet access during training and evaluation.

hugging face

voices: @AndrewCurran_ @ClementDelangue @Gradio @Hesamation @NVIDIAAI @OpenAI @_akhaliq @allen_ai

### 06. [Tesseract creative suite launches with direct integration for Astra AI agents](https://radar.consonance.fyi/en/sujet/829.md)

_agents · 11 independent accounts · 75 posts · 4 labs · 130,984 interactions_

After Effects and Premiere have been rebuilt as Tesseract to provide direct access for GPT-6 Astra and Sol AI agents. The platform connects agents directly to the creative engine to execute design and video workflows.

astra

voices: @AlexFinn @AravSrinivas @OpenAIDevs @aiedge_ @alex_prompter @dair_ai @dotey @emollick

### 07. [Microsoft integrates GPT-6 models and Autopilot into GitHub Copilot](https://radar.consonance.fyi/en/sujet/830.md)

_agents · 3 independent accounts · 14 posts · 2 labs · 37,374 interactions_

Microsoft rolled out a major update to GitHub Copilot, introducing the proactive Autopilot feature alongside two new models: GPT-6 Sol for interactive and agentic coding, and the cost-effective GPT-6 Luna.

github copilot

voices: @AndrewBolis @emollick @mustafasuleyman @rohanpaul_ai @alexeheath @github @jacobandreou @robert_vesely

### 08. [Anthropic releases the Claude Opus 5.5 language model](https://radar.consonance.fyi/en/sujet/831.md)

_architecture · 10 independent accounts · 18 posts · 2 labs · 422,935 interactions_

Anthropic launched Claude Opus 5.5, the first model in its new 5.5 family. It matches the performance of the Fable 5.1 level while reducing running costs by 40% compared to Opus 5.

voices: @AlexFinn @AndrewCurran_ @AnthropicAI @Hesamation @OpenRouter @PromptLLM @aiedge_ @alex_prompter

Same story: [Anthropic releases Claude Opus 5.5 amid tense White House negotiations](https://radar.consonance.fyi/en/sujet/833.md) (12 voices) · [Cursor integrates Claude Opus 5.5 and reduces token costs for agents](https://radar.consonance.fyi/en/sujet/839.md) (8 voices)

### 09. [Muse unveils Muse Charm hardware device and expands agent ecosystem](https://radar.consonance.fyi/en/sujet/834.md)

_agents · 8 independent accounts · 69 posts · 1 articles · 2 labs · 106,966 interactions_

Personal AI agent platform Muse announced partnerships with Shopify, Expedia, and PayPal to handle shopping, travel, and checkout workflows. The company also introduced Muse Charm, a keychain device scheduled to ship in December.

muse spark

voices: @AIatMeta @AlexFinn @aiedge_ @dani_avila7 @emollick @finkd @firstadopter @frankdegods

### 10. [SpaceXAI launches Grok 4.7 model with coding capabilities and low cost](https://radar.consonance.fyi/en/sujet/836.md)

_agents · 5 independent accounts · 87 posts · 1 articles · 1 labs · 2,238,005 interactions_

SpaceXAI released Grok 4.7, scoring 46 on the Artificial Analysis Intelligence Index and outperforming GPT-5.6 Sol on the Coding Agent Index. The release places SpaceXAI third among major AI labs for agentic coding.

grok · grok voice · artificial analysis · deepswe · context window · terminal-bench

voices: @AlexFinn @ArtificialAnlys @CuiMao @aiedge_ @elonmusk @godofprompt @haider1 @minchoi

Same story: [NVIDIA provides accelerated computing infrastructure for SpaceXAI's Grok 4.7](https://radar.consonance.fyi/en/sujet/838.md) (9 voices)

### 11. [Sakana AI introduces Agora-2 world model and appoints Jürgen Schmidhuber](https://radar.consonance.fyi/en/sujet/837.md)

_multimodal · 5 independent accounts · 20 posts · 2 articles · 3 labs · 42,256 interactions_

Sakana AI announced Agora-2, a multi-agent world model supporting up to 20 humans and agents interacting in a real-time shared simulation. Additionally, Jürgen Schmidhuber joined the organization as Chief Scientific Advisor.

world model · david ha

voices: @SchmidhuberAI @_akhaliq @alex_prompter @c_valenzuelab @dair_ai @haider1 @hardmaru @rohanpaul_ai

### 12. [Grok 4.7 xHigh Reaches Competitive Scores on Benchmarks](https://radar.consonance.fyi/en/sujet/841.md)

_evaluation · 7 independent accounts · 57 posts · 2 labs · 122,515 interactions_

Grok 4.7 xHigh reached 58% on Artificial Analysis' AA-Briefcase benchmark, placing just one point behind Claude Fable 5.1 Max. Additionally, a stealth model named Pixel Canary launched with performance matching GPT-6 Astra on Next.js agent evaluations.

kimi k3 · deepseek · glm · deepseek r1

voices: @Hesamation @aiedge_ @alex_prompter @cline @dotey @elonmusk @kimmonismus @rohanpaul_ai

Same story: [Grok 4.7 Improves Scores on Vals Index Following SDK Update](https://radar.consonance.fyi/en/sujet/849.md) (3 voices)

### 13. [Apple and Developers Release New Open-Source Models](https://radar.consonance.fyi/en/sujet/845.md)

_inference · 7 independent accounts · 30 posts · 2 articles · 1 labs · 18,702 interactions_

Apple released a Qwen3.5-9B finetune on Hugging Face designed to convert long documents into page images to save tokens. Other releases include the tev1-4B-experimental classifier and PrismML Bonsai 2 27B, focusing on smaller footprints and local hardware compatibility.

kv cache · mixture of experts · qwen3.8 · qwen3

voices: @AlexFinn @ArtificialAnlys @OpenRouter @PrismML @godofprompt @heyshrutimishra @hwchase17 @rohanpaul_ai

### 14. [Research on Self-Evolving Ontologies Boosts Agent Performance](https://radar.consonance.fyi/en/sujet/848.md)

_agents · 6 independent accounts · 33 posts · 1 articles · 2 labs · 28,109 interactions_

New research demonstrates that implementing an ontology layer for AI agents significantly enhances their capabilities. The study shows that GPT-5.5 gains 26.7 points on DDR-Bench when utilizing self-evolving ontologies.

model context protocol · microsoft research

voices: @ClaudeDevs @aiedge_ @alex_prompter @bcherny @c_valenzuelab @dair_ai @dani_avila7 @dotey

### 15. [DeepSeek Training a 2T Model with Hardware Expansion Plans](https://radar.consonance.fyi/en/sujet/850.md)

_training · 6 independent accounts · 19 posts · 2 labs · 14,934 interactions_

DeepSeek is currently training a 2-trillion-parameter model, with an 8-trillion-parameter model also planned. Additionally, the company expects to receive new Huawei training chips between Q4 2026 and Q1 2027.

deepseek v4.1 flash · deepseek v4 · deepseek v4 flash

voices: @ArtificialAnlys @EMostaque @Hesamation @cline @ollama @op7418 @teortaxesTex @TheTuringPost

### 16. [Google Announces the Googlebook Laptop Lineup](https://radar.consonance.fyi/en/sujet/856.md)

_system_design · 3 independent accounts · 4 posts · 1 labs · 21,230 interactions_

Google has officially introduced a new laptop product line named Googlebook, joining its existing hardware categories to address modern workplace computing needs.

voices: @giffmana @levelsio @sundarpichai @ssamat

### 17. [OpenRouter launches new models and integrations for high-speed decision making](https://radar.consonance.fyi/en/sujet/858.md)

_inference · 3 independent accounts · 37 posts · 1 articles · 14,451 interactions_

OpenRouter has expanded its platform with new model integrations, including Space Bunny Alpha, Typesafe AI's high-speed decision model Jev, and Xiaomi's MiMo-V2.6 series. These additions feature multi-modal capabilities across text, image, video, and audio, alongside 1-million-token context windows.

openrouter

voices: @CompleteSkeptic @OpenRouter @giffmana @heyshrutimishra @KonstantinPilz @alexatallah @chromatique @dasha_shunina

### 18. [Xiaomi releases MiMo-V2.6-Pro achieving top open weights performance](https://radar.consonance.fyi/en/sujet/859.md)

_evaluation · 4 independent accounts · 4 posts · 1 labs · 14,067 interactions_

Xiaomi has introduced MiMo-V2.6-Pro, which debuts as the top open-weights model on the Artificial Analysis Intelligence Index at a cost of $0.13 per task. Built on a classic Grouped Query Attention architecture, the model achieves leading performance in weighted average benchmarks.

voices: @ArtificialAnlys @ClementDelangue @rasbt @teortaxesTex

### 19. [System One models drive new custom agent harnesses and optimization tools](https://radar.consonance.fyi/en/sujet/860.md)

_agents · 3 independent accounts · 11 posts · 1 articles · 1 labs · 5,044 interactions_

The release of System One models like Jev and Contrastive Language Model (CLM) has spurred the development of custom agent harnesses. New optimization tools, such as ReAnchor built with DSPy methodology, leverage these models to enhance verification and execution speed.

dspy

voices: @CompleteSkeptic @lateinteraction @typesafeai @DSPyOSS @MaximeRivest @dbreunig @isaacbmiller1 @omarsar0

### 20. [Google DeepMind publishes new essays on agent swarms and AGI orchestration](https://radar.consonance.fyi/en/sujet/863.md)

_agents · 3 independent accounts · 8 posts · 1 labs · 8,514 interactions_

Google DeepMind has released three new institute essays examining how to control misbehavior in agent swarms, orchestrate complex networks of humans and AIs, and ensure the equitable distribution of AGI benefits.

google deepmind · demis hassabis

voices: @ShaneLegg @hardmaru @mustafasuleyman @Dr_Atoosa @IasonGabriel @blaiseaguera @signulll @tegmark

### 21. [vLLM and SGLang add day-zero support for new Xiaomi and GLM models](https://radar.consonance.fyi/en/sujet/864.md)

_inference · 3 independent accounts · 21 posts · 2 labs · 7,112 interactions_

vLLM has rolled out day-zero support for Xiaomi's MiMo-V2.6 model family, spanning a 1.02T total parameter Pro version and a 309B parameter Flash checkpoint. Additionally, disaggregated setups achieved 469 tok/s single-user decode performance using vLLM on AMD hardware.

vllm · sglang

voices: @AravSrinivas @teortaxesTex @vllm_project @PyTorch @SemiAnalysis_ @XiaomiMiMo @aiDotEngineer @googlegemma

### 22. [New methods improve agent skill evolution and reinforcement learning environments](https://radar.consonance.fyi/en/sujet/873.md)

_training · 5 independent accounts · 12 posts · 2,651 interactions_

Recent research introduces improved methods for evolving agent skills by ranking candidates with a language model, achieving 40% to 70% lower token costs. Work from NVIDIA, such as Skill2Env, further demonstrates compiling public agent skills into reinforcement learning environments.

alibaba qwen · agent skills

voices: @NVIDIAAI @aiedge_ @dair_ai @dani_avila7 @teortaxesTex @tom_doerr @HuggingPapers @claudeskills101

### 23. [Alibaba outlines data center expansion plans as Multilingual TTS Arena launches](https://radar.consonance.fyi/en/sujet/878.md)

_system_design · 3 independent accounts · 12 posts · 1,255 interactions_

Alibaba has targeted 20 gigawatts of global data center capacity by 2032 alongside plans for a 5–10 trillion parameter Qwen model and proprietary AI chips. Simultaneously, the Multilingual TTS Arena leaderboards debuted to compare text-to-speech models across nine languages.

alibaba

voices: @ArtificialAnlys @Hesamation @alex_prompter @teortaxesTex @AlibabaGroup @poezhao0605 @zcode_ai

## Worth reading closely

Nothing yet.

## Hands-on

_Repositories climbing on GitHub today, scored on the day's momentum, rank, adoption, freshness and project health._

- [paperclipai/paperclip](https://github.com/paperclipai/paperclip): An open-source management platform for backend developers to orchestrate and monitor multi-agent workflows in enterprise environments. · +2,109 stars today · ★ 85,508 · TypeScript
- [vectorize-io/hindsight](https://github.com/vectorize-io/hindsight): A long-term memory layer for AI agents that extracts, persists, and learns from historical interactions. · +1,653 stars today · ★ 30,268 · Python
- [mattpocock/skills](https://github.com/mattpocock/skills): A curated collection of prompt engineering skills and shell configurations for engineers optimizing coding agent workflows. · +583 stars today · ★ 269,874 · Shell
- [obra/superpowers](https://github.com/obra/superpowers): A development methodology and skills framework designed to orchestrate software-engineering AI agents. · +468 stars today · ★ 291,766 · Shell
- [dream-num/univer](https://github.com/dream-num/univer): An integrated spreadsheet, doc, and canvas runtime for engineers building agents that need programmatic Office document manipulation. · +1,050 stars today · ★ 18,829 · TypeScript
- [rohitg00/ai-engineering-from-scratch](https://github.com/rohitg00/ai-engineering-from-scratch): A hands-on guide for backend engineers to build AI applications and agents from scratch. · +1,177 stars today · ★ 57,755 · Python
- [google/ax](https://github.com/google/ax): Google's Go-based agentic orchestration runtime for backend engineers building robust, large-scale multi-agent enterprise systems. · +1,379 stars today · ★ 11,661 · Go
- [pbakaus/impeccable](https://github.com/pbakaus/impeccable): A design system specification for developers to guide their coding agents in generating production-grade, aesthetically pleasing user interfaces. · +306 stars today · ★ 71,319 · JavaScript
- [anthropics/skills](https://github.com/anthropics/skills): Official Anthropic toolkit providing reusable skills and capabilities for developers building autonomous agent architectures. · +189 stars today · ★ 178,435 · Python
- [NVIDIA/Model-Optimizer](https://github.com/NVIDIA/Model-Optimizer): NVIDIA's official library for quantizing and optimizing deep learning models to maximize inference throughput. · +359 stars today · ★ 4,549 · Python
- [anthropics/claude-plugins-official](https://github.com/anthropics/claude-plugins-official): An official Anthropic directory of plugins designed to extend the coding capabilities and tool use of Claude Code. · +83 stars today · ★ 36,998 · Python
- [androoAGI/starnet](https://github.com/androoAGI/starnet): A local-first desktop agent harness featuring a pixel-art interface for monitoring multi-agent execution in real time. · +93 stars today · ★ 564 · JavaScript
- [kelseyhightower/kubernetes-the-hard-way](https://github.com/kelseyhightower/kubernetes-the-hard-way): A step-by-step educational guide for provisioning Kubernetes manually, essential for engineers building LLM serving infrastructure. · +119 stars today · ★ 50,166
- [openbao/openbao](https://github.com/openbao/openbao): An open-source secrets management solution to safely store and distribute API keys for backend services and AI pipelines. · +49 stars today · ★ 7,828 · Go

## Next

- [Previous day: 25 September](https://radar.consonance.fyi/en/jour/2026-09-25.md)
- [Next day: 27 September](https://radar.consonance.fyi/en/jour/2026-09-27.md)
- [archives](https://radar.consonance.fyi/en/archives.md)
- [method](https://radar.consonance.fyi/en/methode.md)
- [Consonance](https://radar.consonance.fyi/en/index.md)
- Languages: [Français](https://radar.consonance.fyi/jour/2026-09-26.md) · [Tiếng Việt](https://radar.consonance.fyi/vi/jour/2026-09-26.md)
- For machines: [llms.txt](https://radar.consonance.fyi/llms.txt) · [RSS](https://radar.consonance.fyi/en/rss.xml) · [sitemap.xml](https://radar.consonance.fyi/sitemap.xml)
- HTML version: https://radar.consonance.fyi/en/jour/2026-09-26