Updated
Updated
Showing low-relevance items too. Hide low-relevance items
Thomas Dohmke@ashtomXAI score22AIDeveloper Stefano (@spedemo) built a 3D-printed, remote-controlled robot named Marvin that uses voice detection and speech, with all of it open source. The post says Marvin connects to ChatGPT and Entire, and it will be shown at the WeAreDevs booth 753 in San Jose.
Google Gemini@GeminiAppOfficialAI score12AIGoogle Gemini now integrates with PandaDoc, letting users draft, customize, and send client-ready contracts. Users simply ask Gemini to create a document, such as a Services Agreement for Acme Corp with a $50,000 contract value, using PandaDoc.
Daniel Han@danielhanchenXAI score29AIUnsloth says local AI adoption in 2026 has grown far beyond its own predictions. The company also reports surpassing 500M model downloads on Hugging Face, with Qwen3.8-27B GGUF as its most-downloaded model ever. It says it is working on new features to make local AI faster and more accessible in the coming weeks.
AIComfy Router is now live on the Comfy Developer Platform, giving developers one API to call frontier image, video, 3D, and audio models. Day one models include Seedance 2.5, MiniMax H3, Nano Banana Pro, GPT Image 2, Kling, and Black Forest Labs, and the provider for each job is selectable. Requests fail rather than silently switching providers, and inputs and outputs are deleted after 24 hours.
Why it matters: The post shows how one API key and a provider parameter let developers swap routes for media models without rewriting calls, with failed requests reporting the provider.
Unsloth AI@UnslothAIOfficialAI score22AIUnsloth has passed 500 million model downloads on Hugging Face. Its Qwen3.8-27B GGUF is the most-downloaded model in Unsloth's history.
Mike Knoop@mikeknoopXAI score57AIMike Knoop says the top ARC Prize 2026 ARC-AGI-2 score of 83.06% by Tufa Labs is only 2% short of the 85% grand prize threshold. The challenge runs under strict Kaggle compute limits with no internet access, and the winning solution is set to be open sourced. The image shows the leaderboard with RabbitHole at 76.94%, nvbanana at 74.17%, Yi-Chia Chen at 55.14%, and Kha Vo at 37.50%.
ModelScope@ModelScope2022OfficialAI score40AITeleOCR, a lightweight 1.2B vision-language model released under Apache 2.0, parses digital PDFs and warped phone photos without a separate dewarping model. It scores 96.87 overall on OmniDocBench v1.6, the highest among listed specialized VLMs, and ranks #1 in the ICDAR 2026 Sci-ImageMiner Challenge. It supports structured parsing of text, tables, formulas, layouts, and reading order, with synchronous or asynchronous vLLM inference.
Thomas Dohmke@ashtomXAI score10AIThomas Dohmke met Alex Ksendzovsky at their lab yesterday for a hands-on tour of innovation the lab is developing. The quoted post describes work on neurons meeting silicon, with a partnership with Amazon to bring it to customers and early access available by direct message.
Daniel Han@danielhanchenXAI 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@UnslothAIOfficialAI 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.
Kilo (acq. by Anaconda)@kilocodeOfficialAI score17AIKilo Code tested Grok 4.7 against Opus 5.5 on a touching-grass simulator, with Grok's one-shot result costing 47% less. Per the quoted post, Grok 4.7 cost $3.52 versus $7.35 for Opus 5.5, and both models are available in Kilo now.
Fidji Simo@fidjissimoXAI score46AIOpenAI's Fidji Simo says she is excited to invest in the project and argues education needs a new model. The post offers no further detail, but the quoted background describes The Horowitz Andreessen Academy, a San Francisco school for high school graduates that raised $42M led by a16z.
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.
AICoverage Cat, a licensed insurance brokerage from YC S22, lets users and their AI agents compare umbrella, home, auto, and renters coverage through a portal or Agent API. The company says it sells no leads and earns no commission on the policies its agents recommend, and it is live for shoppers in California, Florida, New York, Texas, and Washington.
LlamaIndex 🦙@llama_indexOfficialAI 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.
Stephanie Palazzolo@steph_palazzoloXAI score23AIAnde, which builds AI agents for corporate entertainment, has raised $52 million in funding led by Lightspeed and Redpoint along with other investors. The round was reported in The Information's AI Agenda newsletter.
WorkBuddy@WorkBuddy_AIOfficialAI score18AIHKUST's Anchor team used WorkBuddy to build two production-ready workbenches and won the Game Track championship. Kaiwu Producer creates a complete FPS game in 8 hours through full-pipeline 3D generation with an AI-driven narrative memory engine and zero human intervention. Anchor is a de-labeling narrative engine that automatically detects stereotypical dependencies.
OpenBMB@OpenBMBOfficialAI 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.
ollama@ollamaOfficialAI score13AIOllama says some requests to the deepseek-v4.1-flash model were charged at an incorrect rate over the last few hours, causing higher-than-expected usage for certain users. Affected accounts have had their monthly or weekly and session usage reset, and any extra usage consumed due to the error has been restored.
OpenBMB@OpenBMBOfficialAI score23AIA developer connected a Raspberry Pi Zero W2 with a webcam and ambient microphone to MiniCPM-o 4.5, an open-source model running locally on a PC. The setup streams audio and video in real time with full-duplex interaction and no cloud APIs.
Kilo (acq. by Anaconda)@kilocodeOfficialAI score36AIKilo's post says Hacktron spent hours failing to exploit a known flaw in an old image library, then a working exploit of OpenAI came within three hours after Claude Opus 5 shipped. The post argues that teams cannot afford model lock-in as frontier models change daily.
AIvllm-metal ports vLLM's scheduler, paged KV cache, and OpenAI-compatible server to Apple Silicon, with MLX and Metal handling execution. The v0.28.0 release added batched MTP, GGUF and hybrid-model support, and faster prefill on M5, and v0.29.0 is installable through Homebrew.
Why it matters: The post explains how vllm-metal packs requests and pages KV cache on Apple Silicon, with benchmarks showing where concurrent serving gains and tradeoffs appear.
LlamaIndex 🦙@llama_indexOfficialAI score22AILlamaIndex has added confidence scores to its Extract product, giving accuracy estimates field by field for extracted data. Developers can use these scores to decide which results their apps accept automatically and which need human review. The feature is available on the Cost Effective, Agentic, and Agentic Plus plans.
LMSYS Org@lmsysorgOfficialPickAI score65AILMSYS Org says NVFP4 KV cache in SGLang fits about 1.78x more context into GPU memory and speeds long-context decoding by up to 78%. Built with Alibaba Qwen and NVIDIA for Blackwell, it stores KV at about 56% of FP8's per-token footprint, with decode throughput up 37%, 58%, and 78% at 32K, 160K, and 1M context. The post reports near-lossless accuracy versus FP8 on GPQA-Diamond and AIME 2025 using Qwen3.5-397B-A17B, and it can be enabled with --kv-cache-dtype nvfp4.
Why it matters: The post gives specific memory and throughput figures for NVFP4 KV cache in SGLang, showing how the format trades cache footprint against long-context decode speed.
WorkBuddy@WorkBuddy_AIOfficialAI score8AIWorkBuddy is tentatively planning a campus visit to MIT on September 24, according to the post. The announcement invites students to attend and bring a friend.
ModelScope@ModelScope2022OfficialAI score3AIModelScope is promoting its booth at Apsara 2026, where visitors can get merchandise such as backpacks, tote bags, mugs, plush pendants, and hats. The booth is located at Booth 1-5C on the first floor of the Intelligence Engine hall at the Hangzhou International Expo Center Phase II.
Elad Gil@eladgilXAI score7AIElad Gil congratulates geom.ai, which is linked to a Business of Home article about Modsy founder Shanna Tellerman returning with a new startup. The post itself gives no details about the startup's product, funding, or focus.
OpenBMB@OpenBMBOfficialAI score23AIDeveloper Mark Fenner built MiniCPM News Desk, a local-first news briefing system powered by MiniCPM5-2B. The model selects key passages from official AI and technology sources, and a rule-based editorial layer preserves dates and context before assembling a daily recap. Invalid or incomplete outputs are rejected, and the full pipeline runs locally without a hosted-model fallback.
AIAridge has signed a cooperation agreement with Ras Al Khaimah to set up a regulatory sandbox for advanced air mobility in the UAE, overseen by the UAE General Civil Aviation Authority. The program will test its Land Aircraft Carrier in desert conditions and plans a demonstration route between Abu Dhabi and Ras Al Khaimah. Regulatory approval for commercial passenger eVTOL flights remains pending.
WorkBuddy@WorkBuddy_AIOfficialAI score12AIWorkBuddy is promoting a feature that lets users pick from their favorite AI models together within the product. Users select a model for each task they want completed.
AIIQAX has surpassed one million electronic Bills of Lading (eBLs), built on the GSBN blockchain and supporting DCSA, BIMCO, and ISO standards. The company is combining AI, IoT, and digital twins to move supply chain management from visibility toward predicting risks, with its AI-powered IoT platform covering more than 200 regions and 11,000 city pairs and about 92,000 connected devices.
OpenBMB@OpenBMBOfficialAI score34AIA research-agent design has GPT-6 orchestrating while MiniCPM5-2B runs locally on DGX Spark to read sources and extract traceable evidence. The post presents this as an example of pairing frontier reasoning with compact, efficient models for transparent research workflows.
OpenBMB@OpenBMBOfficialAI score22AIOpenMed reports that MiniCPM5-2B, run on a local Mac, correctly kept naproxen in medication history rather than the current-medication export after a newer note said it was stopped. Every graph connection in the run links back to its source, using fictional clinical notes.
swyx@swyxXAI score22AIswyx announced a Latent Space podcast episode featuring Jev, subscribable on Apple and YouTube, and thanked guests Allen Park and Ke. A quoted post from @CompleteSkeptic claims Jev is a frontier model with 20-200x faster speed and 40-400x lower cost, but this post itself adds no verified details.
LMSYS Org@lmsysorgOfficialAI score32AIRLinf, an open-source framework for embodied intelligence and AI agents, now supports Cosmos3 from fine-tuning through robot evaluation. With SGLang inference, it delivers 3.33x end-to-end evaluation throughput, batching inference for 128 parallel environments on 8 GPUs across 500 episodes of the full LIBERO-10 evaluation. RLinf also overlaps CPU simulation with GPU inference to reduce waiting between stages.
OpenBMB@OpenBMBOfficialAI score44AIDeveloper @mrgoodmantweets built MiniCPM-o Booking Desk, an open-source appointment booking agent that uses MiniCPM-o 4.5 for real-time, full-duplex voice and audio-visual interaction. The agent listens, speaks, and reads live booking status from an operator screen, while deterministic state control keeps execution reliable. An appointment is only booked after user confirmation.
OpenBMB@OpenBMBOfficialAI score35AIA developer's Augury plant identification model, built on an OpenBMB model, raised photo top-1 accuracy from 71.8% to 80.2% by merging duplicate species keys and adding PCA whitening. The next steps are reaching 90%+ accuracy and building a phone GUI so farmers can use it on-device.
LM Studio@lmstudioOfficialAI score22AIThe Splash engine, posted by LM Studio, is now available as open source on GitHub. The post provides only a link to the incoai/splash repository and includes no further technical details.
LM Studio@lmstudioOfficialAI score20AILM Studio users can enable the Splash engine under Settings > Runtime > Experimental backends and download supported models by searching "incoai." The post recommends an M3 or newer Mac running macOS 26.4 with 36GB+ RAM.
LM Studio@lmstudioOfficialPickAI score62AILM Studio announced that Qwen3.8-27B runs at up to 144 tokens per second on an M5 Max MacBook Pro through its partnership with Inco Splash. The post claims up to 3× the decode speed of Ollama, 2× oMLX, and almost 4× when an agent fans out into sub-agents. Inco Splash is described as an open-source inference engine built for the model and Apple silicon, available through the linked LM Studio blog.
Why it matters: The post reports a concrete decode speed on Apple silicon and compares it against named local inference tools, which helps readers judge local deployment performance.