Updated
Updated
Items with an AI score under 20 are hidden. Show low-relevance items
Lewis Tunstall @ COLM 🌉@_lewtunAI score40AILewis Tunstall of Hugging Face shares his first cameo as a voice actor using the Reachy Mini robot, linked to a YouTube video about Australian football. The post itself provides no technical details, and the Reachy Mini context comes from a separate post by Andi Marafioti about NVIDIA's open-source Nemotron 3 Diarization model.
Julien Chaumond@julien_cAI score36AIHugging Face has released a new JavaScript package, huggingface/lerobot, that lets developers read LeRobot datasets on the Hub directly in the browser without downloading them. The post says coding agents can use it to build custom dataset viewers quickly, with an example UI implemented in about 600 lines of JS.
Qwen@Alibaba_QwenAI score60AIAlibaba's Qwen launched Qwen Intelligence with three mobile agents: a Mobile Planner Agent, a Mobile-Use Agent, and a Mobile Creative Agent. The post reports benchmark results including MobileWorld 82.1, MobileWorld-Real 92.2, and AndroidDaily 97.2, plus a 90% end-to-end success rate, and says the MobilePA-Bench, MobileWorld, MobileWorld-Real, and MobileWorld-Safety benchmarks are open.
Tibor Blaho@btibor91AI score38AIChatGPT can now create interactive flashcards from any topic or from your notes, saved to your Library. Users can flip, shuffle, edit, mark, and read cards aloud, as well as add their own flashcards.
Tencent Hy@TencentHunyuanAI score22AITencent Hunyuan has made its Hy Image3.5 preview available on the OnSolo platform, free for two weeks. The model targets short drama character sheets, full-motion video game assets, and keyframes, with characters kept consistent across episodes and edits that refine rather than regenerate images. OnSolo's background post says the preview supports 5 references at 2K resolution and is free for Members during the two-week window.
WorkBuddy@WorkBuddy_AIAI score36AIWorkBuddy is offering free access to DeepSeek-V4.1-Flash until October 9th. The post invites users to apply the model to their tasks while the offer lasts.
AIPrime Intellect has made Prime Sandboxes generally available, offering each sandbox as a full Linux virtual machine with its own kernel and support for Docker Compose. The product is available through its CLI/SDK and RL suite, with accounts starting at 1,024 concurrent sandboxes, and pricing listed at $0.02 per vCPU-hour, $0.0125 per GiB-hour of memory, and $0.0002 per GiB-hour of disk, valid through December 22. The company says GPU microVMs, snapshotting, sandbox forking, and persistent workspaces are planned next.
Why it matters: The post explains why full VMs rather than gVisor containers matter for agentic RL, since silent environment differences can reward behaviors that fail to transfer.
Tencent Hy@TencentHunyuanAI score43AITencent Hunyuan's Hy Image3.5 preview is now available in ComfyUI, with a claimed 30% higher win rate in human evaluation than Hy Image3.0. The model handles text-to-image and image-to-image in one model at up to 2K resolution, with multilingual text and small print rendering correctly. It also keeps identity and product features consistent across scene, outfit, and style changes.
Daniel Han@danielhanchenAI score22AIUnsloth Desktop received a hotfix update adding image editing for Qwen-Image-2.1. The update also fixes diffusers update issues, GGUF loading failures for Qwen-Image, and black artifacts during diffusion on A100 and consumer GPUs. Users should receive a banner prompting them to update.
Unsloth AI@UnslothAIAI score26AIUnsloth announced that Qwen-Image-2.1 FP8 and GGUF quantized versions should now run properly in Unsloth Desktop. The app supports both image generation and image editing with these quants. Further details are available on the Unsloth GitHub repository.
AIGoogle announced that the Antigravity SDK supports local agent workflows, with initial support for Gemma 4 26B A4B through Google AI Edge's LiteRT. The post includes Python setup steps and says a recommended machine has more than 24GB VRAM or unified memory. It also describes a hybrid pattern in which a cloud Gemini 3.8 Flash planner hands work to local Gemma 4 26B models, with 97.2% of tokens in one recorded run staying local.
Why it matters: The source shows how to run an agent with a local Gemma 4 26B model using LiteRT, plus a hybrid cloud-planner pattern that keeps most tokens on-device.
AIFireworks released ARCv3, a lossless compressor for BF16 weight-update deltas sent from trainers to RL rollout machines. Across 1,000 production RL deltas, ARCv3 produced payloads nearly 50% smaller than ARCv2, averaging about 0.19% of the BF16 weight size versus 0.36%. ARCv3 is available through the Fireworks Training API as fireworks-delta-compression.
AICursor is launching two bots, Rollouts and Security Review, for Teams and Enterprise plans. Rollouts monitors each pull request as it deploys and reports change health per environment, while Security Review reports exploitable bugs on every pull request.
Google Gemma@googlegemmaAI score31AIGoogle Gemma says DiffusionGemma-Jev (djev) can now be deployed as a Jev API-compatible endpoint on Google Cloud Run with a single command. The post reports about 35-60 ms single-step latency and roughly 100-123 requests/sec at batch size 32, at about $3/hr that drops to $0 when idle.
Google Gemma@googlegemmaAI score22AIGoogle Gemma credits @mmastrac and @dylayed for work on DiffusionGemma-Jev (djev), a Jev API-compatible endpoint. Per @dylayed, djev can be deployed to Google Cloud Run with a single gcloud command, at roughly $3/hr while active and $0 when idle.
Grok Bot@botAI score22AIGrok Bot has received recent quality-of-life improvements, including a cleaner interface for working with users' Bots. The post provides no further details on specific features or changes.
Grok Bot@botAI score34AIGrok Bot can now route its internet traffic through the user's own network. The post does not provide further details about how this routing is configured or what limits apply.
Grok Bot@botAI score32AIGrok Bot now natively connects to Google Slides, Sheets, and Docs, according to the post. The update also improves existing integrations, including reading and attaching files to emails.
LiveKit@livekitAI score34AILiveKit has launched Private Links on LiveKit Cloud, giving enterprise agents a fully managed, encrypted tunnel into a customer's VPC. The feature requires no public IPs, inbound firewall rules, or VPN, and is live today in the US and EU.
AISierra says its platform makes enterprise AI agents visible and editable, with journeys, policies, and actions viewable in Agent Studio and testable through Simulations and Experiments before rollout. Customers can export agent logic in a portable structured format, access conversation logs and performance data through export APIs, and manage the agent's code in a Git repository. Sierra agents also connect to existing systems through MCP, REST, GraphQL, or custom integrations.
Grok Bot@botAI score27AIxAI says it rebuilt its customer support around Grok Bot to scale operations without adding headcount. The bot works autonomously to respond to customers, resolve tickets, and manage the support queue.
AIComfyUI's update makes the MiniMax H3 video VAE encode up to about 2.2x faster and decode 1.4-2.7x faster, cutting a 1344x768, 129-frame round trip on an RTX 5090 from 24.3 to 12.7 seconds. The gains come from a fused encoder kernel enabled by default, fp16 accumulation support in a custom convolution, and an int8 decoder, and the source says the changes are visually lossless to the eye. Users need ComfyUI v0.36.0 or above, and the int8 VAE file is a drop-in replacement for the standard one.
Alex Albert@alexalbert__AI score60AIAnthropic's Claude Cowork and chat are being merged into one Claude, which can take a question or report and keep working after the user closes their laptop. When something is unclear, Claude asks, and the user keeps the final say. The update rolls out to Pro and Max over the next few weeks, and the author notes it is coming to everyone soon.
StepFun@StepFun_aiAI score43AIStepFun has open-sourced onPanda, a tool used internally for LLM data annotation and model inspection, letting users correct tokens and let models continue. The company reports a 52% lower median annotation time versus manual post-editing, with SFT and preference data combined in one workflow. It also supports token probability and top-k inspection, token-by-token decoding control, and browser-based testing across SVG generation, web development, and agent tasks.
Waymo@WaymoAI score32AIWaymo has made Teen Accounts officially available in Nashville, giving teens more independence while letting parents feel reassured. The post points readers to for details.
LlamaIndex 🦙@llama_indexAI score22AILlamaIndex released LiteParse v2.14.6, an open-source PDF-to-Markdown parser that processes text-based PDFs about 25% faster. On realistic documents it handled pages at 2.8ms per page, 1.5 times faster than the next-fastest local parser. It runs locally in Python, Node.js, Rust, or directly in the browser.
Mike Krieger@mikeykAI score62AIThe company cut API pricing to $4 per million input tokens and $20 per million output tokens, which it says is 20% less than Opus 5. Cache reads are also 60% cheaper, and subscribers get higher five-hour rate limits plus a banked rate limit reset.
Lydia Hallie ✨@lydiahallieAI score23AIAnthropic's Pro, Max, and Team subscribers can claim a usage limit reset in Settings → Usage, available until Oct 22. Opus 5.5 is now the default for paid plans and is priced lower than Opus 5, so 5-hour and weekly limits go 25% further.
v0@v0AI score52AIv0 announced that users can now use Claude Opus 5.5 in v0, with a link to try it. The quoted Anthropic post says Opus 5.5 is the first model in its new Claude 5.5 family, performs at the level of Claude Fable 5.1 on most tasks, and costs 40% less to run than Opus 5.
AILovable now offers Opus 5.5, which it says matches Opus 5's results while finishing builds in a third to half fewer steps. Internal benchmarks showed Opus 5.5 scoring 4 to 6% ahead of Opus 5 on verification discipline, with step reductions of 26% to 57% and input token reductions of 21% to 59% across tasks.
StepFun@StepFun_aiAI score31AIStepFun released an install script for Step Code that runs on macOS, Linux, or WSL with a single curl command. WSL is the recommended route on Windows, while PowerShell support is currently in beta. The project accepts issues and pull requests on GitHub.
StepFun@StepFun_aiAI score27AIStepFun's Step Code passed 72 of 89 tasks (80.9%) on Terminal-Bench 2.1, tying for the highest pass rate among evaluated harnesses while using fewer tokens than the other tied leaders. On Multi-Frame, it passed 110 of 150 tasks (73.3%) and averaged 5.09M tokens per task, the highest pass rate and lowest token use among six harnesses evaluated.
StepFun@StepFun_aiAI score52AIStepFun has released Step Code v0.1.0, an open-source command-line tool under the MIT License that covers reading and editing code, running tests, and shipping from one CLI. The post reports 80.9% on Terminal-Bench 2.1 and 73.3% on Multi-Frame, a 150-task long-horizon benchmark from StepFun. It also includes one-command static site publishing with StepPage and links the GitHub repository.
Daniel Han@danielhanchenAI score42AIDaniel Han says Qwen-Image-2.1 works in Unsloth Desktop through INT8, FP8, and GGUF builds, with Unsloth also releasing dynamic GGUFs for it. Pinned RAM offloading lets INT8 and FP8 fit under 6–8GB of VRAM while remaining relatively fast. The linked Unsloth post says the 7B model runs on 12GB VRAM and performs on par with Nano Banana 2.0.
Gamma@GammaAppAI score20AIGamma has launched Gamma for Enterprise, a new offering available on its website. The post says it includes features built for organizations that want better presentations, but does not specify which features are included.
Tencent Hy@TencentHunyuanAI score34AITencent Hunyuan has released a preview of Hy Image3.5 on Miora, a design platform, and is offering it free for two weeks. The model keeps existing canvas workflows and remembers users' brand rules while making edits. Miora's background post says the free period runs through October 7 and supports up to 2K output for text-to-image and image-to-image.
Georgi Gerganov@ggerganovAI score60AIHugging Face's transformers library can now load the same GGUF checkpoints used by llama.cpp, with fast local inference on Mac powered by ggml's Metal kernels. The author, Georgi Gerganov, says the work brings those kernels into the transformers ecosystem to increase compatibility and performance.
OpenBMB@OpenBMBAI score59AIOpenBMB highlights VoxWeft, an open-source simultaneous interpretation system for Apple Silicon built by developer @HenryZ30734018 on an MLX implementation of VoxCPM2. The system turns live speech into translated speech on-device, with first audio streaming in about 170 ms on an M5 MacBook. VoxCPM2 generates speech in 30 languages, supports direct language-pair interpretation, and clones a target voice from about 5 seconds of reference audio.
AIT-Head, Alibaba's chip subsidiary, unveiled the Zhenwu V900 AI chip for training and inference at the 2026 Apsara Conference in Hangzhou. The company claims three times the performance of its predecessor, the Zhenwu M890, with 216GB of memory, 1,200GB/s inter-chip bandwidth, and mass production expected in the first quarter of 2027.
Kimi.ai@Kimi_MoonshotAI score46AIKimi has released its Kimi Browser Extension, formerly Kimi WebBridge, which runs in the browser sidebar to navigate websites and fill out forms. Users can record repetitive steps once and save them as a skill for Kimi to reuse later. The extension is available now on the Chrome Web Store.
