The Hermes Agent Creative Hackathon sponsored by @Kimi_Moonshot has ended! Finalists were selected by Nous and Kimi staff out of 227 submissions on creativity, usefulness and presentation. We were absolutely blown away by the creativity of the things you all built using Hermes. It was tough choosing winners, huge thank you to all who participated! Winners below:

Second place: Ambien by @ya1sec A daydreaming agent running inside Hermes Agent that samples a curated Obsidian corpus for resonant idea collisions and synthesizes them into published essays, leveraging the harness for iterative voice calibration that keeps the prose from reading like an LLM. https://t.co/NqjjPuGHAg

Third place: Hermes Agent Teletype emulator by @hughpyle A 1960s Teletype ASR33 brought back online by Hermes Agent running on a Raspberry Pi behind it, with a plugin emulating the machine's Ambisonic-recorded hum and keystrokes, collapsing the telegraph age into the agent age. https://t.co/PfR5LAMs6V

Kimi track - First place: JumpFoundry by @evvaaannnn A Hermes Agent plugin where the designer hand-draws one or two glyphs and Hermes Agent generates the rest, acting as coordinator that routes per-glyph revisions through parallel model calls before exporting an editable TTF. https://t.co/SYapiI8fwP

Kimi track - Second place: sacred_not_secret by @macbethAI An audio-reactive architectural wireframing piece where Hermes Agent sourced and optimized the temple STLs and built the custom GLSL control app with sliders and a waveform timeline. https://t.co/vAyIzufuqV https://t.co/yLBF0h04sl

Kimi track - Third place: Permanence by @MachYear A personal taste-keeping workspace where Hermes Agent synthesizes your references with you and reshapes its own interface to match your aesthetics. https://t.co/vKMYA1qb43 https://t.co/TinkL3i63S

To add some clarity: you don't pay extra. It's the same subscription, same price per month. What's new our sub now covers two separate pools: · Interactive → sub limits, unchanged · Programmatic → new $20–$200 included(!!) credit, metered at API rates https://t.co/knrcT1P5Of

Here I am, reading Homer’s Odyssey directly from the Greek text and thinking about how much Christopher Nolan disrespected us. Don’t pay for the movie. Resist this slowly dying agenda that is rotting our societies. Hollywood is dead. https://t.co/f9hLbUTAbi

Langship is not @langchain. It's an open-source alternative TO langchain's deployment + governance stack. Self-hostable. apache 2.0. There's a full manifesto on the site. someone really said hold my beer #langship #langchain https://t.co/BrMsdvSMav

+2 more

Reaching 2M impressions on X! Thanks everyone for keeping it real against AI hype! Last 3 months have been a blast. Answering rumours: • No, I’m not an AI. • Realist: neither pro nor anti. • AI research: mention me on any paper for my analysis Trailblazing the manifold. https://t.co/aiRUvyn3ya

🪄 Your agent experience, refined. The latest @code release brings better BYOK visibility and control, integrated browser improvements, and more. It also introduces the new Agents window (preview), making it easier to explore, iterate on, and review tasks across multiple projects.

More updates in @code: Markdown preview for diffs, terminal risk assessment and output compression, inline plan editing for Claude and Copilot CLI, and more. 🔗 Explore the weekly updates: https://t.co/Jiq3E46KIZ https://t.co/lUOVRdH5HX

Run powerful, self-improving AI agents from your desk. @NousResearch's Hermes Agent brings reliable, self-evolving agentic AI to NVIDIA RTX PCs and DGX Spark. Get started. 👇 https://t.co/BBJNCnBzsu

Just spent 8 days across Beijing, Shanghai, and Hangzhou meeting with the labs actually building Chinese AI. 5 major open model releases dropped during the week I was there. What I came away with isn't what the policy debate would have you expect. 🧵 https://t.co/RSHJ0HQYFI https://t.co/Au8DWfMuBE

CausalCine Real-Time Autoregressive Generation for Multi-Shot Video Narratives https://t.co/WjFiFd6Y8J

paper: https://t.co/EigvXfnSCp

Apple presents Unmasking On-Policy Distillation Where It Helps, Where It Hurts, and Why https://t.co/wKR7JR705R

paper: https://t.co/I8ccIFfep3

Streaming an Android phone to my Mac in a data center via Tailscale + https://t.co/iT7Eq2zYG7 and my claw controls it via https://t.co/2cXk37Lxt7. Now my claw can order me an Uber. https://t.co/q3xXgy2bTY

Fun Fact - Over 50% of French enterprises (CAC40) are on @huggingface To secure AI usage in organizations: https://t.co/XH3XKkeslk https://t.co/YATU517Gmt

@MLCatttt I usually look at the config.json and eyeball it based on other architectures (that I already have in the https://t.co/T63kGUgsrp)

AI video creation is evolving FAST 🚀 Just saw this workflow using Agent One inside InVideo, and honestly, the cinematic consistency is impressive. You describe the idea, and AI helps with: • Scenes • Camera angles • Editing • Workflow consistency This space is moving incredibly fast right now 👀 Excited to see where AI agents + creative workflows will go next. #AI #AIVideo #AItools #GenerativeAI #ContentCreation #AIagents #InVideo

Very cool blog post by @mishig25 on the state of local agents The @ArtificialAnlys intelligence index increases faster than Moore's Law! https://t.co/yAHSUcOUl5

Time-lapse of a Scarlet Pimpernel flower opening - makes me happy to look over from my desk and see these little flowers doing their thing 😀😀 https://t.co/uUEuvxvhok

Here, a neural cellular automata has learned to guess what is being (badly) drawn - watch how the correct consensus prediction emerges over time. Each pixel 'cell' can only see its neighbors, and each runs the same simple NN. https://t.co/OYx2MElsiE

Today we release Token Superposition Training (TST), a modification to the standard LLM pretraining loop that produces a 2-3× wall-clock speedup at matched FLOPs without changing the model architecture, optimizer, tokenizer, or training data. During the first third of training, the model reads and predicts contiguous bags of tokens, averaging their embeddings on the input side and predicting the next bag with a modified cross-entropy on the output side. For the remainder of the run, it trains normally on next-token prediction. The inference-time model is identical to one produced by conventional pretraining. Validated at 270M, 600M, and 3B dense scales, and at 10B-A1B MoE. The work on TST was led by @bloc97_, @gigant_theo, and @theemozilla.

Do you know which of your Hermes cron jobs are consuming your cost and token budget? Observe. Measure. Optimize. Repo in thread 👇 https://t.co/eVbXfEKkZt

PyTorch 2.12 introduces major updates across compilation, export, distributed training, and accelerator support. Highlights include up to 100x faster batched linalg.eigh on CUDA, the new torch.accelerator.Graph API, Microscaling quantization support in torch .export.save, and fused Adagrad. The release includes 2,926 commits from 457 contributors since PyTorch 2.11. Have questions? Join @AndreyTalman (@Meta), @albanDesmaison (@Meta), and @joespeez (@reflection_ai), moderated by @Chris_AI_HPC (@Meta), on May 20 at 10:00 AM PT for a live Q&A covering the release and answering questions from the community. 🔗 Read the release blog and register for the webinar: https://t.co/lSkHPD3FQR #PyTorch #OpenSourceAI #MachineLearning #AIInfrastructure

Poolside is hosting a 2-day model research hackathon in London. Join us to push an open-weight agent model as far as you can. RL and fine-tune Laguna XS.2, our latest-generation model, on Prime Intellect Lab. Dates: May 29–30 Partners: @nvidia + @PrimeIntellect + @huggingface Prize: NVIDIA DGX Spark Agents need better models. Better models need cracked researchers. Link below.

Founders building outside of NYC or SF: https://t.co/0vNLYmWXt6

Your Agent Can Now Train Models The argument from @mervenoyann: open source models have caught up. GLM 5.1 is leading the Artificial Analysis intelligence index over closed models, and the gap is closing with every release cycle. Weight access means you can quantize, fine tune, and deploy to edge devices without data leaving your infrastructure. https://t.co/kQvBd0uHuk The talk covers the Hugging Face ecosystem built for agentic work: inference providers with tool use routing, benchmark datasets for filtering by SWE bench scores on Hub, a traces repository type for storing agent sessions, and skills that plug into coding agents. The closer is a live demo: she asks Claude Code to fine tune a vision language model on a dataset by name. The agent calculates VRAM requirements, picks an instance, and kicks off the job. What used to be a day of napkin math is now a prompt.

Want to (officially) use Codex at work? Send this post to your CTO to bring your team to Codex. Eligible enterprise customers who switch in the next 30 days get 2 free months of Codex usage for new users. https://t.co/38e8y7MAmg