---
title: "Consonance · CARD NO. 2026-09-24"
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/2026-09-24"
lang: "en"
published: "2026-09-24"
updated: "2026-09-24T07:43:24.738Z"
---

# Consonance · CARD NO. 2026-09-24

> 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. [OpenAI releases GPT-6 Sol and GPT-6 Luna alongside Tesseract video tools](https://radar.consonance.fyi/en/sujet/683.md)

_multimodal · 15 independent accounts · 80 posts · 1 articles · 8 labs · 480,254 interactions_

OpenAI has introduced the GPT-6 Sol and GPT-6 Luna models, building on the capabilities of GPT-6 Astra to offer faster and more affordable options. Additionally, the company launched Tesseract, a video creative suite designed to give AI agents direct access to video editing tools.

gpt-6

voices: @AndrewCurran_ @ArtificialAnlys @Hesamation @OpenAI @OpenAIDevs @OpenRouter @PromptLLM @aiedge_

Same story: [OpenAI rolls out GPT-6 Sol and Luna broadly alongside a 50% API price cut](https://radar.consonance.fyi/en/sujet/689.md) (3 voices) · [Astra launches Astra for Law alongside GPT-6 Sol and Luna models](https://radar.consonance.fyi/en/sujet/705.md) (14 voices)

### 02. [SpaceXAI launches Grok 4.7 with high performance on the Artificial Analysis Intelligence Index](https://radar.consonance.fyi/en/sujet/684.md)

_evaluation · 8 independent accounts · 116 posts · 1 articles · 1 labs · 2,424,577 interactions_

SpaceXAI has released Grok 4.7, which scores 46 on the Artificial Analysis Intelligence Index and places the lab in the top four AI labs. The model demonstrates improved coding agent performance while maintaining high speed and low cost.

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

voices: @AlexFinn @AndrewCurran_ @ArtificialAnlys @GithubProjects @Hesamation @aiedge_ @dotey @elonmusk

### 03. [Anthropic releases Claude Opus 5.5 with 40% lower operating costs](https://radar.consonance.fyi/en/sujet/685.md)

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

Anthropic has introduced Claude Opus 5.5, the first model in the Claude 5.5 family, delivering performance comparable to Claude Fable 5.1 while cutting operating costs by 40% compared to Opus 5.

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

Same story: [Claude Opus 5.5 arrives in Cursor and the Claude Marketplace](https://radar.consonance.fyi/en/sujet/688.md) (8 voices) · [Anthropic releases Claude Opus 5.5 featuring optimized token efficiency and performance](https://radar.consonance.fyi/en/sujet/692.md) (10 voices)

### 04. [DigitalOcean announces public preview of Managed Agents](https://radar.consonance.fyi/en/sujet/690.md)

_system_design · 9 independent accounts · 60 posts · 3 labs · 180,825 interactions_

DigitalOcean has launched a public preview of Managed Agents, allowing users to run tools like Claude Code or Codex within runtime environments that pause automatically during idle periods.

codex

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

### 05. [NVIDIA provides accelerated computing infrastructure for Grok 4.7 and the AI ecosystem](https://radar.consonance.fyi/en/sujet/691.md)

_system_design · 10 independent accounts · 48 posts · 2 labs · 190,994 interactions_

NVIDIA's accelerated computing infrastructure is supporting models across the AI ecosystem, including powering SpaceXAI's Grok 4.7 for demanding coding and knowledge workflows.

nvidia

voices: @ArtificialAnlys @CuiMao @NVIDIAAI @SchmidhuberAI @alex_prompter @dair_ai @elonmusk @firstadopter

### 06. [ChatGPT expands voice capabilities, multi-account plugins, and Chrome extensions](https://radar.consonance.fyi/en/sujet/693.md)

_agents · 8 independent accounts · 66 posts · 2 articles · 4 labs · 106,885 interactions_

OpenAI updated ChatGPT voice capabilities to integrate plugins across email, calendar, and Slack. The platform also added support for linking multiple accounts with plugins and running Chrome extensions inside the desktop app.

chatgpt

voices: @AndrewBolis @CuiMao @OpenAI @OpenAIDevs @VibeMarketer_ @alex_prompter @dotey @gdb

### 07. [Google launches Gemini 3.8 Flash and Flash-Lite TTS audio model family](https://radar.consonance.fyi/en/sujet/696.md)

_multimodal · 10 independent accounts · 32 posts · 2 articles · 3 labs · 51,900 interactions_

Google introduced Gemini 3.8 Flash TTS and Gemini 3.8 Flash-Lite TTS featuring over 2,000 production-ready voices and support for 100 languages. The release also extends real-time audio capabilities through Gemini 3.8 Live in search.

gemini 3.8 · big bench audio

voices: @Alibaba_Qwen @AndrewCurran_ @ArtificialAnlys @EMostaque @Google @GoogleAIStudio @OfficialLoganK @aiedge_

### 08. [DeepSeek advances large model development and outlines hardware training roadmap](https://radar.consonance.fyi/en/sujet/697.md)

_architecture · 6 independent accounts · 28 posts · 3 labs · 121,304 interactions_

DeepSeek is currently training 2-trillion and 8-trillion parameter models, with plans to incorporate Huawei training chips in late 2026 or early 2027. The platform also integrated the DeepSeek V4.1 Flash model into developer tools.

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

voices: @EMostaque @Hesamation @elonmusk @ollama @op7418 @teortaxesTex @vllm_project @TheAhmadOsman

### 09. [Muse launches Mac app integrating versatile AI agent assistant for developers](https://radar.consonance.fyi/en/sujet/698.md)

_agents · 10 independent accounts · 74 posts · 1 articles · 2 labs · 215,272 interactions_

The Muse app for Mac has been released, operating across files, calendar, notes, and messages. The platform also opened API connector access for developers and partnered with Shopify to streamline shopping workflows.

muse spark · meta ai

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

### 10. [Google introduces Googlebook laptop line powered by Android and ChromeOS stack](https://radar.consonance.fyi/en/sujet/700.md)

_system_design · 9 independent accounts · 73 posts · 1 articles · 2 labs · 56,190 interactions_

Google launched Googlebook, a new laptop featuring a 2.8K OLED touchscreen display and a 14-hour battery life. The device is built on the Android technology stack paired with ChromeOS desktop foundations.

gemini · gemini 3.1 flash

voices: @Alibaba_Qwen @AndrewCurran_ @ArtificialAnlys @GeminiApp @GithubProjects @Google @GoogleDeepMind @Hesamation

Same story: [Google introduces an entirely new laptop product line called Googlebook](https://radar.consonance.fyi/en/sujet/715.md) (3 voices)

### 11. [Anthropic partners with Accenture on independent evaluation and establishes biological lab](https://radar.consonance.fyi/en/sujet/701.md)

_evaluation · 11 independent accounts · 78 posts · 1 articles · 3 labs · 104,057 interactions_

Anthropic partnered with Accenture for independent evaluations of frontier AI systems, committing at least $1 billion to building capacity. The company has also established a physical wet lab in San Francisco for biological research.

anthropic · claude sonnet · claude haiku · aime

voices: @AlexFinn @AndrewCurran_ @AnthropicAI @CuiMao @Hesamation @aiedge_ @alex_prompter @dair_ai

### 12. [NVIDIA releases Nemotron 3 Diarization model for multi-speaker audio tracking](https://radar.consonance.fyi/en/sujet/702.md)

_multimodal · 6 independent accounts · 43 posts · 9 articles · 4 labs · 101,287 interactions_

NVIDIA released the Nemotron 3 Diarization model with 100 million parameters, capable of tracking up to eight speakers and handling overlapping voices. The system identifies who spoke and when within multi-person audio transcripts.

chain of thought · hugging face

voices: @AndrewCurran_ @ClementDelangue @GithubProjects @Gradio @Hesamation @NVIDIAAI @_akhaliq @alex_prompter

### 13. [Claude Code adds AGENTS.md configuration support and cloud session capabilities](https://radar.consonance.fyi/en/sujet/703.md)

_agents · 11 independent accounts · 69 posts · 2 articles · 1 labs · 226,932 interactions_

Claude Code updated to version 2.1.277, adding support for AGENTS.md configuration files when CLAUDE.md is absent. Cloud sessions also exited research preview, enabling Claude to run parallel threads even after laptops are closed.

claude code

voices: @AlexFinn @ArtificialAnlys @ClaudeDevs @CuiMao @aiedge_ @alex_prompter @alliekmiller @bcherny

### 14. [Introduction of the CC agent and lobbying efforts over AI regulation](https://radar.consonance.fyi/en/sujet/706.md)

_agents · 12 independent accounts · 45 posts · 1 articles · 3 labs · 109,756 interactions_

Google vừa công bố CC, một trợ lý AI được thiết kế để hỗ trợ các gia đình quản lý lịch trình và công việc chung. Trong một diễn biến khác, các lãnh đạo công nghệ lớn được cho là đã thúc ép chính quyền loại bỏ một đề xuất quản lý AI do người đứng đầu Google DeepMind đề xuất.

google · google deepmind · demis hassabis · jason wei

voices: @Google @Hesamation @JeffDean @OfficialLoganK @aiedge_ @alliekmiller @boringmarketer @dair_ai

### 15. [Research on self-evolving ontologies and MCP integration for agents](https://radar.consonance.fyi/en/sujet/709.md)

_agents · 7 independent accounts · 29 posts · 1 articles · 3 labs · 46,961 interactions_

Recent research demonstrates that incorporating a self-evolving ontology layer significantly improves agent performance on complex benchmarks. Meanwhile, the ecosystem continues to expand with new integrations and servers built around the Model Context Protocol.

model context protocol · microsoft research

voices: @aiedge_ @alex_prompter @c_valenzuelab @dair_ai @dani_avila7 @dotey @ideabrowser @op7418

### 16. [PixVerse unveils the R2 world model and the impact of harness design](https://radar.consonance.fyi/en/sujet/710.md)

_multimodal · 3 independent accounts · 11 posts · 1 articles · 3 labs · 25,064 interactions_

PixVerse has released R2, a real-time world model that allows users to interact with and edit dynamic environments using prompts. Additionally, recent studies highlight that employing an explicit graph harness can drastically improve a model's performance on long-horizon robotic tasks without altering the underlying weights.

yann lecun · world model

voices: @_akhaliq @alex_prompter @c_valenzuelab @dair_ai @tszzl @10ayaGoldsen @AmOptimistShow @FuturLucide

### 17. [Release of Ternary Bonsai 2 27B and the Qwen3.8 real-time translation model](https://radar.consonance.fyi/en/sujet/712.md)

_inference · 8 independent accounts · 52 posts · 2 articles · 3 labs · 120,411 interactions_

Ternary Bonsai 2 27B has been released, utilizing a 9x reduction in size compared to its Qwen3.8 base model while retaining 98.2% of its aggregate benchmark performance. Additionally, the Qwen3.8 ecosystem expanded with LiveTranslate, a simultaneous interpretation model built on an interleave architecture.

kv cache · mixture of experts · qwen3.8 · qwen3 · tokens per second

voices: @Alibaba_Qwen @ArtificialAnlys @EMostaque @OpenRouter @PrismML @godofprompt @heyshrutimishra @hwchase17

### 18. [MiMo-V2.6-Pro leads the open weights rankings on Artificial Analysis](https://radar.consonance.fyi/en/sujet/717.md)

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

MiMo-V2.6-Pro has debuted as the top open weights model on the Artificial Analysis Intelligence Index. Built on a classic grouped query attention architecture, it establishes a strong position on the intelligence versus cost per task Pareto frontier.

voices: @ArtificialAnlys @ClementDelangue @rasbt @teortaxesTex

### 19. [Grok 4.7 xHigh competes at the top alongside updates on Kimi K3](https://radar.consonance.fyi/en/sujet/718.md)

_evaluation · 7 independent accounts · 34 posts · 1 labs · 19,598 interactions_

Grok 4.7 xHigh has climbed close to the top spot on Artificial Analysis' AA-Briefcase benchmark, trailing the leader by a single point. Meanwhile, indicators point toward an upcoming Kimi K3.1 release, alongside demonstrations of running large models locally on standard CPUs.

glm · kimi k3

voices: @AndrewCurran_ @Hesamation @OpenRouter @aiedge_ @cline @kimmonismus @nutlope @rohanpaul_ai

### 20. [Discussions on automated AI research and multi-agent systems](https://radar.consonance.fyi/en/sujet/720.md)

_agents · 3 independent accounts · 15 posts · 1 labs · 37,455 interactions_

The community discussed the potential of automating scientific research using multi-agent systems. Conversations centered on the prospects and implications of superintelligence driving autonomous breakthroughs in mathematics and engineering.

noam brown

voices: @deanwball @haider1 @jerryjliu0 @polynoamial @scaling01 @teortaxesTex @tszzl @dwarkesh_sp

### 21. [Optimizing inference systems and accelerating execution for large language models](https://radar.consonance.fyi/en/sujet/722.md)

_inference · 3 independent accounts · 23 posts · 1 articles · 2 labs · 9,227 interactions_

Zhipu reported that GLM-5.3-Flash runs across more than 10,000 domestic AI accelerators, achieving a 3.2x increase in throughput through automated optimization driven by an AI agent. Meanwhile, models like DiffusionGemma and Ternary-Bonsai received updates within vLLM and MLX frameworks.

sglang · vllm

voices: @AndrewCurran_ @AravSrinivas @dotey @teortaxesTex @vllm_project @PyTorch @XiaomiMiMo @aiDotEngineer

### 22. [DeepSeek trials large-scale server infrastructure and Xiaomi launches MiMo-V2.6](https://radar.consonance.fyi/en/sujet/725.md)

_architecture · 3 independent accounts · 48 posts · 2 labs · 8,396 interactions_

DeepSeek operates production-scale units spanning around 160 nodes, serving about 3 million sandboxes per day. Simultaneously, Xiaomi released the MiMo-V2.6 series featuring Pro and Flash models, with the Pro variant scoring 46 points on Artificial Analysis.

deepseek · deepseek r1 · moonshot ai

voices: @Hesamation @MatthewBerman @dotey @kimmonismus @teortaxesTex @54158zxc @PeterDiamandis @SemiAnalysis_

### 23. [Running local open-weight models on constrained hardware devices](https://radar.consonance.fyi/en/sujet/726.md)

_inference · 4 independent accounts · 20 posts · 3 labs · 11,277 interactions_

Developers deployed a 27-billion-parameter ternary model at 1.72 bits per weight on a single graphics card with 12GB of VRAM. Additionally, updates brought llama.cpp's Metal kernels to transformers and added SGLang support for image generation models.

llama.cpp · llama · quantization · unsloth · gemma

voices: @Alibaba_Qwen @emollick @teortaxesTex @vllm_project @ylecun @DogukanUrker @HuggingModels @NewsFromGoogle

### 24. [Alibaba announces large-scale AI infrastructure plans and RecreationWorld sandbox](https://radar.consonance.fyi/en/sujet/737.md)

_system_design · 3 independent accounts · 15 posts · 1 labs · 1,804 interactions_

Alibaba targeted 20 gigawatts of global data center capacity by 2032 alongside plans for a 5 to 10 trillion parameter Qwen model series. The Qwen team also released RecreationWorld, a five-platform sandbox environment for computer-use agents.

bytedance · alibaba · alibaba qwen

voices: @AravSrinivas @Hesamation @alex_prompter @rohanpaul_ai @teortaxesTex @AlibabaGroup @HuggingPapers @perplexity_ai

### 25. [Improving skill optimization methods for AI agents](https://radar.consonance.fyi/en/sujet/739.md)

_agents · 3 independent accounts · 9 posts · 1,980 interactions_

Researchers introduced a method allowing AI agents to improve skill prompts by ranking candidates, achieving 40% to 70% lower token costs compared to conventional methods. NVIDIA also detailed an approach to compile public agent skills into reinforcement learning environments.

agent skills

voices: @NVIDIAAI @aiedge_ @dair_ai @dani_avila7 @claudeskills101

### 26. [Evaluating safety refusal rates on coding agent benchmarks](https://radar.consonance.fyi/en/sujet/744.md)

_evaluation · 3 independent accounts · 5 posts · 1,583 interactions_

Artificial Analysis introduced safety refusal reporting in version 1.5 of its Coding Agent Index to help explain model behavioral differences. Data showed distinct variations in safety refusal rates among evaluated models during coding tasks.

devin

voices: @ArtificialAnlys @Hesamation @NVIDIAAI @0x15f

### 27. [Automated learning applications and improved retrieval methods for finance](https://radar.consonance.fyi/en/sujet/745.md)

_agents · 3 independent accounts · 9 posts · 1,393 interactions_

A joint university study introduced a path for financial agents to learn from SEC filing errors by requiring all new behaviors to pass regression tests first. New adaptive retrieval frameworks were also released to better handle specific data queries.

retrieval augmented generation

voices: @AndrewBolis @GithubProjects @alex_prompter @rohanpaul_ai @boltzbit

## Worth reading closely

Nothing yet.

## Hands-on

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

- [anthropics/financial-services](https://github.com/anthropics/financial-services): A collection of LLM integration examples for backend engineers building secure financial applications with Claude. · +664 stars today · ★ 37,102 · Python
- [davila7/claude-code-templates](https://github.com/davila7/claude-code-templates): A CLI utility for configuring and monitoring Claude Code, designed for developers optimizing their terminal-based AI coding workflows. · +389 stars today · ★ 31,625 · 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. · +1,543 stars today · ★ 9,465 · Go
- [obra/superpowers](https://github.com/obra/superpowers): A development methodology and skills framework designed to orchestrate software-engineering AI agents. · +474 stars today · ★ 290,870 · 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. · +1,142 stars today · ★ 16,544 · TypeScript
- [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. · +304 stars today · ★ 70,514 · JavaScript
- [agent-substrate/substrate](https://github.com/agent-substrate/substrate): A Go-based core systems infrastructure for backend engineers building scalable execution environments for autonomous AI agents. · +558 stars today · ★ 3,618 · Go
- [superdesigndev/treg](https://github.com/superdesigndev/treg): An OpenRouter-style proxy specifically for agent tool calls, helping backend engineers route and manage LLM tool executions. · +506 stars today · ★ 2,861 · Python
- [browser-use/video-use](https://github.com/browser-use/video-use): A coding agent framework for automated video editing, built for engineers developing LLM-driven multimedia processing pipelines. · +746 stars today · ★ 26,647 · Python
- [BuilderIO/agent-native](https://github.com/BuilderIO/agent-native): A TypeScript framework that helps backend engineers design and deploy complex agentic applications. · +87 stars today · ★ 6,678 · TypeScript
- [DeusData/codebase-memory-mcp](https://github.com/DeusData/codebase-memory-mcp): High-performance code intelligence MCP server. Indexes codebases into a persistent knowledge graph — average repo in milliseconds. 158 languages, sub-ms queries, 99% fewer tokens. Single static binary, zero dependencies. · +190 stars today · ★ 44,714 · C
- [strands-agents/harness-sdk](https://github.com/strands-agents/harness-sdk): An open-source SDK for building and orchestrating production-grade agent harnesses in Python and TypeScript. · +115 stars today · ★ 7,982 · Python
- [HKUDS/CLI-Anything](https://github.com/HKUDS/CLI-Anything): A framework that wraps standard software into agent-native CLI interfaces for seamless automation. · +57 stars today · ★ 50,065 · Python
- [TNT-Likely/PanWatch](https://github.com/TNT-Likely/PanWatch): 盯盘侠 PanWatch · 自托管 AI 盯盘助手，集成 TradingAgents 多 Agent 投资决策 \| A股/港股/美股实时监控、持仓管理、智能分析、全渠道推送 · +95 stars today · ★ 1,666 · Python
- [harry7557558/spirula-studio](https://github.com/harry7557558/spirula-studio): Cross-vendor 3D Gaussian Splatting trainer - video to splat to mesh, Vulkan or CUDA. · +69 stars today · ★ 824 · C++

## Claims

_Checkable assertions with their sources and their contradictions. Each stands on at least two independent sources, or a person read it first; the stamp says which._

### 01. Claude discovered a previously uncharacterized enzyme system with features reminiscent of CRISPR after AI agents searched through a DNA sequence database.

_TWO SOURCES · NOT REREAD · 24 September 2026 · claude · 2 sources_

Condition: With 950 agents running over 21 hours.

- **evidence** · [@nc_frey](https://x.com/nc_frey/status/2102831099969687847): “We’ve set up a molecular biology lab at Anthropic and we’re announcing our first discovery! Claude discovered a new CRISPR-like enzyme. 950 agents spent 21 hours searching through a database of DNA sequences until one of the agents found something striking”
- **evidence** · [@BoWang87](https://x.com/BoWang87/status/2102874527487369693): “About 950 Claude agent sessions ran over 21 hours, surveyed ~200,000 reverse transcriptases, and produced 19 reports.”

**Still to check**
- Re-examine the surveyed DNA database and experimentally verify the enzyme's activity in the lab.

## Next

- [Previous day: 23 September](https://radar.consonance.fyi/en/jour/2026-09-23.md)
- [Next day: 25 September](https://radar.consonance.fyi/en/jour/2026-09-25.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/2026-09-24.md) · [Tiếng Việt](https://radar.consonance.fyi/vi/jour/2026-09-24.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/2026-09-24