{"schema_version":"onlylabs.public_analysis.v1","url":"https://onlylabs.fyi/analysis/nous","json_url":"https://onlylabs.fyi/analysis/nous/analysis.json","evidence_json_url":"https://onlylabs.fyi/analysis/nous/evidence.json","generated_at":"2026-06-28T02:16:54.220Z","analysis":{"org_slug":"nous","url":"https://onlylabs.fyi/analysis/nous","json_url":"https://onlylabs.fyi/analysis/nous/analysis.json","evidence_json_url":"https://onlylabs.fyi/analysis/nous/evidence.json","dossier_url":"https://onlylabs.fyi/labs/nous","org":{"slug":"nous","name":"Nous Research","category":"neolab","category_label":"Neolab","homepage_url":"https://nousresearch.com/"},"title":"Nous Research analysis","summary":"Nous Research is in the midst of a rapid productization pivot. Hermes Agent — a 190k-star open-source AI agent platform — is receiving weekly releases with ~1,475 commits per cycle and 245 community contributors. Simultaneously, the org is shipping frontier models (Hermes-4 family through 405B scale), building RL training infrastructure (Atropos at 1,273 stars), and advancing distributed training research…","markdown":"## Thesis\n\nNous Research is in the midst of a rapid productization pivot. Hermes Agent — a 190k-star open-source AI agent platform — is receiving **weekly releases** with ~1,475 commits per cycle and 245 community contributors [P1](https://github.com/NousResearch/hermes-agent/releases/tag/v2026.6.19)[P24](https://github.com/NousResearch/hermes-agent). Simultaneously, the org is shipping frontier models (Hermes-4 family through 405B scale), building RL training infrastructure (Atropos at 1,273 stars), and advancing distributed training research (DisTrO/Psyche Network). A cluster of **May 2026 NVIDIA infrastructure forks** (Megatron-LM, Megatron-Bridge, NeMo RL, NeMo Gym, Nemotron) alongside a General Counsel hire and Forward Deployed Engineer role signal a lab moving from research collective toward platform company with enterprise posture [E19](https://nousresearch.com/careers)[E20](https://nousresearch.com/careers)[E22](https://nousresearch.com/careers)[E29](https://github.com/NousResearch/Megatron-LM)[E30](https://github.com/NousResearch/Megatron-Bridge)[E59](https://github.com/NousResearch/RL)[E60](https://github.com/NousResearch/Gym).\n\n## Signal desks\n\n### Hiring\n\n- **UI/UX Designer** — open role implies consumer-facing product buildout for Hermes Desktop and agent interfaces [E19](https://nousresearch.com/careers).\n- **General Counsel** — first in-house legal hire, signals platform maturity, regulatory posture, and/or commercial contracting needs [E20](https://nousresearch.com/careers).\n- **Full Stack Engineer** — product engineering for web/desktop surfaces around Hermes Agent [E21](https://nousresearch.com/careers).\n- **Forward Deployed Engineer** — enterprise/partner deployment role, suggests organizational customers onboarding onto Hermes or Forge infrastructure [E22](https://nousresearch.com/careers).\n- **Research Scientist** — continued fundamental research capacity, likely spanning RL (Atropos), long-context attention, and reasoning [E23](https://nousresearch.com/careers).\n- **Machine Learning Engineer** — training and fine-tuning pipeline scaling for the Hermes-4 family, MoE models, and Atropos RL training [E24](https://nousresearch.com/careers).\n\n### Forks\n\n- **NVIDIA training/infra cluster** — Nous forked NVIDIA's Megatron-LM, Megatron-Bridge, Automodel, NeMo RL, NeMo Gym, and Nemotron in late April–May 2026, suggesting large-scale training infrastructure buildout or inspection for Hermes model family and Atropos RL workloads [E28](https://github.com/NousResearch/Automodel)[E29](https://github.com/NousResearch/Megatron-LM)[E30](https://github.com/NousResearch/Megatron-Bridge)[E46](https://github.com/NousResearch/Nemotron)[E59](https://github.com/NousResearch/RL)[E60](https://github.com/NousResearch/Gym).\n- **NVIDIA desktop/shell tools** — OpenShell, OpenShell-Community, and NemoClaw forks (April 2026) point to Hermes Desktop terminal and sandbox capabilities under active development [E47](https://github.com/NousResearch/NemoClaw)[E51](https://github.com/NousResearch/OpenShell)[E52](https://github.com/NousResearch/OpenShell-Community).\n- **Microsoft agent-governance-toolkit** — fork (May 2026, 31 stars) signals interest in enterprise agent governance, compliance, and controlled-agent deployment patterns [E37](https://github.com/NousResearch/agent-governance-toolkit).\n- **vercel-labs/wterm** and **vadimdemedes/ink** — terminal-UI tools forked in April 2026, consistent with Hermes Desktop surface development [E50](https://github.com/NousResearch/wterm)[E58](https://github.com/NousResearch/ink).\n\n### Releases\n\n- **Hermes Agent v0.17.0 (v2026.6.19)** — \"The Reach Release\": iMessage channel via Photon, Raft agent network gateway, upgraded subagents, image generation editing, Cursor Composer via xAI Grok, dashboard profile builder, major Skills Hub rework, `memory` tool overhaul, 300+ issues closed [P1](https://github.com/NousResearch/hermes-agent/releases/tag/v2026.6.19)[E16](https://github.com/NousResearch/hermes-agent/releases/tag/v2026.6.19).\n- **Hermes Agent v0.16.0 (v2026.6.5)** — \"The Surface Release\": Hermes Desktop public preview on Windows/macOS/Linux; NVIDIA/skills added as built-in trusted skills tap alongside OpenAI, Anthropic, HuggingFace [W2](http://github.com/NousResearch/hermes-agent/releases/tag/v2026.6.5)[E18](https://github.com/NousResearch/hermes-agent/releases/tag/v2026.6.5).\n- **Hermes Agent weekly cadence** — versions v2026.5.29.2, v2026.5.29, v2026.5.28, v2026.5.16, v2026.5.7, v2026.4.30, v2026.4.23, v2026.4.16, v2026.4.13 all shipped between mid-April and early June 2026 [E25](https://github.com/NousResearch/hermes-agent/releases/tag/v2026.5.29.2)[E26](https://github.com/NousResearch/hermes-agent/releases/tag/v2026.5.29)[E27](https://github.com/NousResearch/hermes-agent/releases/tag/v2026.5.28)[E35](https://github.com/NousResearch/hermes-agent/releases/tag/v2026.5.16)[E41](https://github.com/NousResearch/hermes-agent/releases/tag/v2026.5.7)[E44](https://github.com/NousResearch/hermes-agent/releases/tag/v2026.4.30)[E48](https://github.com/NousResearch/hermes-agent/releases/tag/v2026.4.23)[E53](https://github.com/NousResearch/hermes-agent/releases/tag/v2026.4.16)[E57](https://github.com/NousResearch/hermes-agent/releases/tag/v2026.4.13).\n- **Model: Hermes-4-405B** (Aug 2025) — 405B parameter hybrid-reasoning chat model, llama3 license, 286 downloads [E15](https://huggingface.co/NousResearch/Hermes-4-405B).\n- **Model: Hermes-4-70B** (Aug 2025) — 70B frontier chat model, 4,062 downloads [E8](https://huggingface.co/NousResearch/Hermes-4-70B).\n- **Model: Hermes-4-14B** (Aug 2025) — 14B model, Apache-2.0 license, 31,365 downloads [E11](https://huggingface.co/NousResearch/Hermes-4-14B).\n- **Model: Hermes-4.3-36B** (Nov 2025) — 36B, Apache-2.0, 25,308 downloads [E6](https://huggingface.co/NousResearch/Hermes-4.3-36B).\n- **Model: NousCoder-14B** (Dec 2025) — code-focused model, Apache-2.0, 212 likes [E7](https://huggingface.co/NousResearch/NousCoder-14B).\n- **Model: nomos-1** (Dec 2025) — 30.5B reasoning model for mathematical proof-writing, Apache-2.0 [E12](https://huggingface.co/NousResearch/nomos-1)[P25](https://github.com/NousResearch/nomos).\n- **Model: Minos-v1** (Apr 2025) — 396M text-classification model (likely a reward/judge model), Apache-2.0 [E10](https://huggingface.co/NousResearch/Minos-v1).\n- **Model: moe-10b-a1b-8k-wsd-lr3e4-1t** (Mar 2026) — experimental MoE architecture, 10B total / ~1B active parameters [E45](https://huggingface.co/NousResearch/moe-10b-a1b-8k-wsd-lr3e4-1t).\n- **DeepHermes Atropos specialist models** — ToolCalling-Specialist, Financial-Fundamentals-Prediction-Specialist, and AscensionMaze-RLAIF variants, all 8B Llama-3-based, pipeline tagged `reinforcement-learning` [E39](https://huggingface.co/NousResearch/DeepHermes-ToolCalling-Specialist-Atropos)[E43](https://huggingface.co/NousResearch/DeepHermes-Financial-Fundamentals-Prediction-Specialist-Atropos)[E54](https://huggingface.co/NousResearch/DeepHermes-AscensionMaze-RLAIF-8b-Atropos).\n- **Atropos RL Framework** (v2025.4.29+) — 1,273 stars, 367 forks, LLM RL gym for trajectory collection and evaluation, active as of June 2026 [P21](https://github.com/NousResearch/atropos)[E33](https://github.com/NousResearch/atropos).\n- **tinker-atropos** — integration layer connecting Atropos with Thinking Machines Tinker API for managed RL training [P26](https://github.com/NousResearch/tinker-atropos).\n- **Hermes Agent self-evolution** (Mar 2026) — 4,364-star repo: DSPy + GEPA framework for optimizing agent skills, prompts, and code [E36](https://github.com/NousResearch/hermes-agent-self-evolution).\n- **Hermes paperclip-adapter** (Mar 2026) — 1,650-star TypeScript adapter to run Hermes as a managed employee in Paperclip companies [E55](https://github.com/NousResearch/hermes-paperclip-adapter).\n- **hermes-compression-eval** (May 2026) — offline probe-based evaluation harness for Hermes Agent's ContextCompressor, adapted from Factory's compression evaluation methodology [E34](https://github.com/NousResearch/hermes-compression-eval).\n- **kanban-video-pipeline** (May 2026) — 43-star new repo, likely a media generation pipeline [E42](https://github.com/NousResearch/kanban-video-pipeline).\n\n### Talking\n\n- **Lighthouse Attention** — Nous proposed and open-sourced a selection-based hierarchical attention mechanism delivering 1.4–1.7× pretraining speedup at 98K context and ~17× faster than standard attention at 512K on a single B200. Covered by MarkTechPost and Newsference in mid-May 2026 [W3](https://www.marktechpost.com/2026/05/16/nous-research-proposes-lighthouse-attention-a-training-only-selection-based-hierarchical-attention-that-delivers-1-4-1-7x-pretraining-speedup-at-long-context/)[W4](https://newsference.com/c/nous-lighthouse-attention-long-context-speedup).\n- **Hermes Desktop launch** — The Decoder covered the Hermes Desktop public preview (MIT license, Windows/macOS/Linux), calling it \"an open-source AI agent for every platform,\" highlighting sub-agents, sandboxed execution (local/Docker/SSH/Singularity/Modal), and access to 300+ models via Nous Portal [W6](https://the-decoder.com/nous-research-releases-hermes-desktop-an-open-source-ai-agent-for-every-platform/).\n- **Nous Portal integration** — teknium1 authored a dedicated Portal integration page, positioning Hermes-4-70B and Hermes-4-405B as \"frontier hybrid-reasoning chat models\" available at heavily discounted rates, alongside Google Gemini, Anthropic Claude, and other providers through OpenRouter-style aggregation [W1](https://github.com/NousResearch/hermes-agent/commit/2666009ccc00f653a1db4ff249c9b1aabdb34f13).\n- **GitHub org overview** — as of late May 2026, NousResearch's top repos by stars: hermes-agent (200k+), hermes-agent-self-evolution (4.3k), hermes-paperclip-adapter (1.6k), Atropos (1.3k+), plus autonovel experimental writing pipeline [W5](https://github.com/nousresearch).\n\n## Shipping\n\nThe pace of shipping is aggressive and multi-vector. **Hermes Agent is on a weekly release cadence** — at least 8 versioned releases between April 13 and June 19, 2026, with v0.17.0 alone representing ~1,475 commits, ~800 merged PRs, and 245 community contributors [P1](https://github.com/NousResearch/hermes-agent/releases/tag/v2026.6.19)[E16](https://github.com/NousResearch/hermes-agent/releases/tag/v2026.6.19)[E25](https://github.com/NousResearch/hermes-agent/releases/tag/v2026.5.29.2)[E26](https://github.com/NousResearch/hermes-agent/releases/tag/v2026.5.29)[E27](https://github.com/NousResearch/hermes-agent/releases/tag/v2026.5.28)[E35](https://github.com/NousResearch/hermes-agent/releases/tag/v2026.5.16)[E41](https://github.com/NousResearch/hermes-agent/releases/tag/v2026.5.7)[E44](https://github.com/NousResearch/hermes-agent/releases/tag/v2026.4.30)[E48](https://github.com/NousResearch/hermes-agent/releases/tag/v2026.4.23)[E53](https://github.com/NousResearch/hermes-agent/releases/tag/v2026.4.16)[E57](https://github.com/NousResearch/hermes-agent/releases/tag/v2026.4.13). The agent has expanded from a local tool to a multi-channel platform (iMessage, Raft agent network, Telegram, desktop) with five sandboxed execution backends [P1](https://github.com/NousResearch/hermes-agent/releases/tag/v2026.6.19)[W6](https://the-decoder.com/nous-research-releases-hermes-desktop-an-open-source-ai-agent-for-every-platform/). **Model releases span the parameter spectrum**: 405B down to 1B, with specialized variants for tool-calling, financial prediction, and code (NousCoder-14B) [E7](https://huggingface.co/NousResearch/NousCoder-14B)[E15](https://huggingface.co/NousResearch/Hermes-4-405B)[E39](https://huggingface.co/NousResearch/DeepHermes-ToolCalling-Specialist-Atropos)[E40](https://huggingface.co/NousResearch/Llama-3.2-1B)[E43](https://huggingface.co/NousResearch/DeepHermes-Financial-Fundamentals-Prediction-Specialist-Atropos). The **Atropos RL framework** shipped alongside tinker-atropos for managed training, indicating a buildout pipeline where environments → RL training → specialist model → deployment is being systematized [P21](https://github.com/NousResearch/atropos)[P26](https://github.com/NousResearch/tinker-atropos)[E33](https://github.com/NousResearch/atropos).\n\n## Research themes\n\n1. **Long-context efficiency** — Lighthouse Attention demonstrates a selection-based hierarchical attention for pretraining, targeting 1.4–1.7× speedup at scale without custom sparse kernels or auxiliary losses [W3](https://www.marktechpost.com/2026/05/16/nous-research-proposes-lighthouse-attention-a-training-only-selection-based-hierarchical-attention-that-delivers-1-4-1-7x-pretraining-speedup-at-long-context/)[W4](https://newsference.com/c/nous-lighthouse-attention-long-context-speedup). The `scaling-transformer` repo (Jul 2025) and `infini-attention` repo (Jul 2024) suggest sustained attention-architecture exploration [P8](https://github.com/NousResearch/infini-attention)[P20](https://github.com/NousResearch/scaling-transformer).\n2. **Reinforcement learning for LLMs** — Atropos is a dedicated LLM RL gym, and the DeepHermes Atropos specialist models (ToolCalling, Financial, AscensionMaze-RLAIF) show an end-to-end RL training pipeline producing task-specialized models [P21](https://github.com/NousResearch/atropos)[E39](https://huggingface.co/NousResearch/DeepHermes-ToolCalling-Specialist-Atropos)[E43](https://huggingface.co/NousResearch/DeepHermes-Financial-Fundamentals-Prediction-Specialist-Atropos)[E54](https://huggingface.co/NousResearch/DeepHermes-AscensionMaze-RLAIF-8b-Atropos). The NeMo RL and NeMo Gym forks (Apr 2026) indicate deeper RL infrastructure investment [E59](https://github.com/NousResearch/RL)[E60](https://github.com/NousResearch/Gym).\n3. **Distributed training** — DisTrO/Psyche Network claims 3–4 orders of magnitude reduction in inter-GPU communication, demonstrated by training a Consilience 40B model [P10](https://github.com/NousResearch/DisTrO). Production code lives in the PsycheFoundation org in Rust (`distro.rs`) [P10](https://github.com/NousResearch/DisTrO).\n4. **Reasoning and math** — Open-Reasoning-Tasks (493 stars) curates reasoning task collections; Nomos (194 stars) is a reasoning harness for mathematical proof-writing with self-critique loops; Forge API offered reasoning completions at fast/medium/slow presets [P6](https://github.com/NousResearch/Open-Reasoning-Tasks)[P14](https://github.com/NousResearch/forge-api-demo)[P25](https://github.com/NousResearch/nomos).\n5. **Agent self-improvement** — hermes-agent-self-evolution (4,364 stars) uses DSPy + GEPA to optimize agent skills, prompts, and code autonomously [E36](https://github.com/NousResearch/hermes-agent-self-evolution).\n6. **MoE architectures** — The `moe-10b-a1b-8k-wsd-lr3e4-1t` model (10B total, ~1B active) shows experimentation with mixture-of-experts at modest scale [E45](https://huggingface.co/NousResearch/moe-10b-a1b-8k-wsd-lr3e4-1t).\n7. **Alternative-language tooling** — Kaida (Kotlin LLM library), wandb-rs (Rust Weights & Biases client), solana-flake (Nix/Solana build tooling), and iroh-fake-store (Rust blob store testing) indicate polyglot infrastructure investment [P11](https://github.com/NousResearch/solana-flake)[P12](https://github.com/NousResearch/wandb-rs)[P17](https://github.com/NousResearch/kaida)[P27](https://github.com/NousResearch/iroh-fake-store).\n\n## Hiring & scaling\n\nSix open roles as of June 2026 paint a picture of productization + enterprise maturity: **UI/UX Designer** and **Full Stack Engineer** for consumer-grade Hermes Desktop surfaces; **Forward Deployed Engineer** for enterprise customer onboarding; **General Counsel** signaling legal/commercial readiness for platform-scale operations; **Research Scientist** and **Machine Learning Engineer** sustaining model R&D [E19](https://nousresearch.com/careers)[E20](https://nousresearch.com/careers)[E21](https://nousresearch.com/careers)[E22](https://nousresearch.com/careers)[E23](https://nousresearch.com/careers)[E24](https://nousresearch.com/careers). No cited evidence on team sizes, locations, or compensation bands — the job listings lack descriptions beyond titles and the ATS source marker.\n\nThe fork pattern strongly implies a **NVIDIA partnership or heavy NeMo/Megatron dependency**: 9 of the 12 cited forks (May 2026 cluster) are NVIDIA repos covering the full training stack — Megatron-LM (distributed training), Megatron-Bridge (NeMo adapter), Automodel (model selection), NeMo RL/Gym (RL infrastructure), Nemotron (model architecture), OpenShell/OpenShell-Community (terminal), NemoClaw (agent shell tools) [E28](https://github.com/NousResearch/Automodel)[E29](https://github.com/NousResearch/Megatron-LM)[E30](https://github.com/NousResearch/Megatron-Bridge)[E37](https://github.com/NousResearch/agent-governance-toolkit)[E46](https://github.com/NousResearch/Nemotron)[E47](https://github.com/NousResearch/NemoClaw)[E50](https://github.com/NousResearch/wterm)[E51](https://github.com/NousResearch/OpenShell)[E52](https://github.com/NousResearch/OpenShell-Community)[E58](https://github.com/NousResearch/ink)[E59](https://github.com/NousResearch/RL)[E60](https://github.com/NousResearch/Gym). Combined with NVIDIA/skills becoming a built-in trusted skills tap in Hermes Agent [W2](http://github.com/NousResearch/hermes-agent/releases/tag/v2026.6.5), this suggests a deepening technical integration.\n\n## Category implications\n\n**Infrastructure** — The Megatron-LM + Megatron-Bridge + NeMo RL forks indicate Nous is building or inspecting large-scale training infrastructure on NVIDIA's stack. Combined with DisTrO's distributed training claims and the Psyche Network, Nous is investing in training infrastructure that could reduce dependency on centralized compute providers [E29](https://github.com/NousResearch/Megatron-LM)[E30](https://github.com/NousResearch/Megatron-Bridge)[E59](https://github.com/NousResearch/RL)[E60](https://github.com/NousResearch/Gym)[P10](https://github.com/NousResearch/DisTrO). The wandb-rs Rust client and iroh-fake-store suggest in-house MLOps tooling in Rust [P12](https://github.com/NousResearch/wandb-rs)[P27](https://github.com/NousResearch/iroh-fake-store).\n\n**Product** — Hermes Agent is the primary product vector: a multi-channel (desktop, iMessage, Telegram, Raft), multi-model (300+ via Portal), self-improving agent platform with 190k+ GitHub stars [P1](https://github.com/NousResearch/hermes-agent/releases/tag/v2026.6.19)[P24](https://github.com/NousResearch/hermes-agent)[W6](https://the-decoder.com/nous-research-releases-hermes-desktop-an-open-source-ai-agent-for-every-platform/). The paperclip-adapter (\"run Hermes as a managed employee in a Paperclip company\") and agent-governance-toolkit fork point toward enterprise agent deployment and governance as a GTM motion [E37](https://github.com/NousResearch/agent-governance-toolkit)[E55](https://github.com/NousResearch/hermes-paperclip-adapter).\n\n**Research** — RL is the dominant research theme, with Atropos as gym, specialist models as outputs, and thinking-machines Tinker integration for managed training [P21](https://github.com/NousResearch/atropos)[P26](https://github.com/NousResearch/tinker-atropos)[E39](https://huggingface.co/NousResearch/DeepHermes-ToolCalling-Specialist-Atropos)[E43](https://huggingface.co/NousResearch/DeepHermes-Financial-Fundamentals-Prediction-Specialist-Atropos)[E54](https://huggingface.co/NousResearch/DeepHermes-AscensionMaze-RLAIF-8b-Atropos). Long-context attention (Lighthouse) and MoE architectures represent efficiency research bets [W3](https://www.marktechpost.com/2026/05/16/nous-research-proposes-lighthouse-attention-a-training-only-selection-based-hierarchical-attention-that-delivers-1-4-1-7x-pretraining-speedup-at-long-context/)[W4](https://newsference.com/c/nous-lighthouse-attention-long-context-speedup)[E45](https://huggingface.co/NousResearch/moe-10b-a1b-8k-wsd-lr3e4-1t). Open-Reasoning-Tasks and Nomos anchor the reasoning research track [P6](https://github.com/NousResearch/Open-Reasoning-Tasks)[P25](https://github.com/NousResearch/nomos).\n\n**Hiring** — The mix of product roles (UI/UX, Full Stack, Forward Deployed) alongside research roles (Research Scientist, MLE) and a General Counsel indicates a transition from pure research lab to product company with enterprise sales motion [E19](https://nousresearch.com/careers)[E20](https://nousresearch.com/careers)[E21](https://nousresearch.com/careers)[E22](https://nousresearch.com/careers)[E23](https://nousresearch.com/careers)[E24](https://nousresearch.com/careers).\n\n**GTM** — The Nous Portal aggregates 300+ models at discounted rates, positioning Hermes-4 models as a funnel into the platform [W1](https://github.com/NousResearch/hermes-agent/commit/2666009ccc00f653a1db4ff249c9b1aabdb34f13). Hermes Desktop's free MIT-licensed public preview serves as a top-of-funnel acquisition channel. The Forward Deployed Engineer role signals enterprise deployment deals [E22](https://nousresearch.com/careers). The General Counsel hire indicates contracts, compliance, and possibly revenue-generation maturity [E20](https://nousresearch.com/careers). No cited evidence of revenue figures, pricing tiers beyond \"heavily discounted rates,\" or paying customer counts.\n\n## Traction highlights\n\n- **hermes-agent**: 190,109 stars, 32,944 forks, 19,913 open issues — top-tier open-source traction rivaling major frameworks [P24](https://github.com/NousResearch/hermes-agent). The HN event at repo creation scored only 4 points/3 comments [E1](https://github.com/NousResearch/hermes-agent), but subsequent growth is extraordinary.\n- **hermes-agent-self-evolution**: 4,364 stars since March 2026 [E36](https://github.com/NousResearch/hermes-agent-self-evolution).\n- **hermes-paperclip-adapter**: 1,650 stars since March 2026 [E55](https://github.com/NousResearch/hermes-paperclip-adapter).\n- **atropos**: 1,273 stars, 367 forks since April 2025 [P21](https://github.com/NousResearch/atropos)[E33](https://github.com/NousResearch/atropos).\n- **DisTrO**: 1,034 stars, 88 points/26 comments on HN at launch [P10](https://github.com/NousResearch/DisTrO)[E3](https://github.com/NousResearch/DisTrO).\n- **Hermes-Function-Calling**: 1,386 stars, 198 forks [P9](https://github.com/NousResearch/Hermes-Function-Calling).\n- **Open-Reasoning-Tasks**: 493 stars, 72 forks [P6](https://github.com/NousResearch/Open-Reasoning-Tasks).\n- **Hermes-3-Llama-3.1-8B**: 312,589 HuggingFace downloads, 464 likes [E2](https://huggingface.co/NousResearch/Hermes-3-Llama-3.1-8B).\n- **Hermes-3-Llama-3.1-70B**: 131,870 downloads [E13](https://huggingface.co/NousResearch/Hermes-3-Llama-3.1-70B).\n- **Hermes-4-14B**: 31,365 downloads [E11](https://huggingface.co/NousResearch/Hermes-4-14B).\n- **Hermes-4.3-36B**: 25,308 downloads [E6](https://huggingface.co/NousResearch/Hermes-4.3-36B).\n\nCommunity contribution volume is notable: v0.17.0 alone had 245 contributors and 300+ issues closed [P1](https://github.com/NousResearch/hermes-agent/releases/tag/v2026.6.19).\n\n## Sources\n\n[P1](https://github.com/NousResearch/hermes-agent/releases/tag/v2026.6.19) hermes-agent v2026.6.19 release notes\n[P6](https://github.com/NousResearch/Open-Reasoning-Tasks) Open-Reasoning-Tasks repo\n[P8](https://github.com/NousResearch/infini-attention) infini-attention repo\n[P9](https://github.com/NousResearch/Hermes-Function-Calling) Hermes-Function-Calling repo\n[P10](https://github.com/NousResearch/DisTrO) DisTrO repo\n[P11](https://github.com/NousResearch/solana-flake) solana-flake repo\n[P12](https://github.com/NousResearch/wandb-rs) wandb-rs repo\n[P14](https://github.com/NousResearch/forge-api-demo) forge-api-demo repo\n[P17](https://github.com/NousResearch/kaida) kaida repo\n[P20](https://github.com/NousResearch/scaling-transformer) scaling-transformer repo\n[P21](https://github.com/NousResearch/atropos) atropos repo\n[P24](https://github.com/NousResearch/hermes-agent) hermes-agent repo metadata\n[P25](https://github.com/NousResearch/nomos) nomos repo\n[P26](https://github.com/NousResearch/tinker-atropos) tinker-atropos repo\n[P27](https://github.com/NousResearch/iroh-fake-store) iroh-fake-store repo\n[E1](https://github.com/NousResearch/hermes-agent) hermes-agent repo creation event\n[E2](https://huggingface.co/NousResearch/Hermes-3-Llama-3.1-8B) Hermes-3-Llama-3.1-8B model release\n[E3](https://github.com/NousResearch/DisTrO) DisTrO repo creation event\n[E6](https://huggingface.co/NousResearch/Hermes-4.3-36B) Hermes-4.3-36B model release\n[E7](https://huggingface.co/NousResearch/NousCoder-14B) NousCoder-14B model release\n[E8](https://huggingface.co/NousResearch/Hermes-4-70B) Hermes-4-70B model release\n[E10](https://huggingface.co/NousResearch/Minos-v1) Minos-v1 model release\n[E11](https://huggingface.co/NousResearch/Hermes-4-14B) Hermes-4-14B model release\n[E12](https://huggingface.co/NousResearch/nomos-1) nomos-1 model release\n[E15](https://huggingface.co/NousResearch/Hermes-4-405B) Hermes-4-405B model release\n[E16](https://github.com/NousResearch/hermes-agent/releases/tag/v2026.6.19) hermes-agent v2026.6.19 release event\n[E18](https://github.com/NousResearch/hermes-agent/releases/tag/v2026.6.5) hermes-agent v2026.6.5 release event\n[E19](https://nousresearch.com/careers) UI/UX Designer job opening\n[E20](https://nousresearch.com/careers) General Counsel job opening\n[E21](https://nousresearch.com/careers) Full Stack Engineer job opening\n[E22](https://nousresearch.com/careers) Forward Deployed Engineer job opening\n[E23](https://nousresearch.com/careers) Research Scientist job opening\n[E24](https://nousresearch.com/careers) Machine Learning Engineer job opening\n[E25](https://github.com/NousResearch/hermes-agent/releases/tag/v2026.5.29.2) hermes-agent v2026.5.29.2 release\n[E26](https://github.com/NousResearch/hermes-agent/releases/tag/v2026.5.29) hermes-agent v2026.5.29 release\n[E27](https://github.com/NousResearch/hermes-agent/releases/tag/v2026.5.28) hermes-agent v2026.5.28 release\n[E28](https://github.com/NousResearch/Automodel) Automodel fork event\n[E29](https://github.com/NousResearch/Megatron-LM) Megatron-LM fork event\n[E30](https://github.com/NousResearch/Megatron-Bridge) Megatron-Bridge fork event\n[E33](https://github.com/NousResearch/atropos) atropos repo creation event\n[E34](https://github.com/NousResearch/hermes-compression-eval) hermes-compression-eval repo creation\n[E35](https://github.com/NousResearch/hermes-agent/releases/tag/v2026.5.16) hermes-agent v2026.5.16 release\n[E36](https://github.com/NousResearch/hermes-agent-self-evolution) hermes-agent-self-evolution repo creation\n[E37](https://github.com/NousResearch/agent-governance-toolkit) agent-governance-toolkit fork event\n[E39](https://huggingface.co/NousResearch/DeepHermes-ToolCalling-Specialist-Atropos) DeepHermes-ToolCalling-Specialist-Atropos model release\n[E40](https://huggingface.co/NousResearch/Llama-3.2-1B) Llama-3.2-1B model release\n[E41](https://github.com/NousResearch/hermes-agent/releases/tag/v2026.5.7) hermes-agent v2026.5.7 release\n[E42](https://github.com/NousResearch/kanban-video-pipeline) kanban-video-pipeline repo creation\n[E43](https://huggingface.co/NousResearch/DeepHermes-Financial-Fundamentals-Prediction-Specialist-Atropos) DeepHermes-Financial-Fundamentals-Prediction-Specialist-Atropos model release\n[E44](https://github.com/NousResearch/hermes-agent/releases/tag/v2026.4.30) hermes-agent v2026.4.30 release\n[E45](https://huggingface.co/NousResearch/moe-10b-a1b-8k-wsd-lr3e4-1t) moe-10b-a1b-8k-wsd-lr3e4-1t model release\n[E46](https://github.com/NousResearch/Nemotron) Nemotron fork event\n[E47](https://github.com/NousResearch/NemoClaw) NemoClaw fork event\n[E48](https://github.com/NousResearch/hermes-agent/releases/tag/v2026.4.23) hermes-agent v2026.4.23 release\n[E50](https://github.com/NousResearch/wterm) wterm fork event\n[E51](https://github.com/NousResearch/OpenShell) OpenShell fork event\n[E52](https://github.com/NousResearch/OpenShell-Community) OpenShell-Community fork event\n[E53](https://github.com/NousResearch/hermes-agent/releases/tag/v2026.4.16) hermes-agent v2026.4.16 release\n[E54](https://huggingface.co/NousResearch/DeepHermes-AscensionMaze-RLAIF-8b-Atropos) DeepHermes-AscensionMaze-RLAIF-8b-Atropos model release\n[E55](https://github.com/NousResearch/hermes-paperclip-adapter) hermes-paperclip-adapter repo creation\n[E57](https://github.com/NousResearch/hermes-agent/releases/tag/v2026.4.13) hermes-agent v2026.4.13 release\n[E58](https://github.com/NousResearch/ink) ink fork event\n[E59](https://github.com/NousResearch/RL) RL fork event (NVIDIA-NeMo)\n[E60](https://github.com/NousResearch/Gym) Gym fork event (NVIDIA-NeMo)\n[W1](https://github.com/NousResearch/hermes-agent/commit/2666009ccc00f653a1db4ff249c9b1aabdb34f13) Nous Portal integration commit\n[W2](http://github.com/NousResearch/hermes-agent/releases/tag/v2026.6.5) Hermes Agent v0.16.0 release page\n[W3](https://www.marktechpost.com/2026/05/16/nous-research-proposes-lighthouse-attention-a-training-only-selection-based-hierarchical-attention-that-delivers-1-4-1-7x-pretraining-speedup-at-long-context/) MarkTechPost Lighthouse Attention coverage\n[W4](https://newsference.com/c/nous-lighthouse-attention-long-context-speedup) Newsference Lighthouse Attention coverage\n[W5](https://github.com/nousresearch) Nous Research GitHub org overview\n[W6](https://the-decoder.com/nous-research-releases-hermes-desktop-an-open-source-ai-agent-for-every-platform/) The Decoder Hermes Desktop coverage","generated_at":"2026-06-27T19:34:24.815+00:00","citations":[{"url":"https://github.com/NousResearch/hermes-agent/releases/tag/v2026.6.19","path":null,"label":"NousResearch/hermes-agent","type":"external"},{"url":"https://github.com/NousResearch/Open-Reasoning-Tasks","path":null,"label":"NousResearch/Open-Reasoning-Tasks","type":"external"},{"url":"https://github.com/NousResearch/infini-attention","path":null,"label":"NousResearch/infini-attention","type":"external"},{"url":"https://github.com/NousResearch/Hermes-Function-Calling","path":null,"label":"NousResearch/Hermes-Function-Calling","type":"external"},{"url":"https://github.com/NousResearch/DisTrO","path":null,"label":"NousResearch/DisTrO","type":"external"},{"url":"https://github.com/NousResearch/solana-flake","path":null,"label":"NousResearch/solana-flake","type":"external"},{"url":"https://github.com/NousResearch/wandb-rs","path":null,"label":"NousResearch/wandb-rs","type":"external"},{"url":"https://github.com/NousResearch/forge-api-demo","path":null,"label":"NousResearch/forge-api-demo","type":"external"},{"url":"https://github.com/NousResearch/kaida","path":null,"label":"NousResearch/kaida","type":"external"},{"url":"https://github.com/NousResearch/scaling-transformer","path":null,"label":"NousResearch/scaling-transformer","type":"external"},{"url":"https://github.com/NousResearch/atropos","path":null,"label":"NousResearch/atropos","type":"external"},{"url":"https://github.com/NousResearch/hermes-agent","path":null,"label":"NousResearch/hermes-agent","type":"external"},{"url":"https://github.com/NousResearch/nomos","path":null,"label":"NousResearch/nomos","type":"external"},{"url":"https://github.com/NousResearch/tinker-atropos","path":null,"label":"NousResearch/tinker-atropos","type":"external"},{"url":"https://github.com/NousResearch/iroh-fake-store","path":null,"label":"NousResearch/iroh-fake-store","type":"external"},{"url":"https://github.com/NousResearch/hermes-agent","path":null,"label":"NousResearch/hermes-agent","type":"external"},{"url":"https://huggingface.co/NousResearch/Hermes-3-Llama-3.1-8B","path":null,"label":"NousResearch/Hermes-3-Llama-3.1-8B","type":"external"},{"url":"https://github.com/NousResearch/DisTrO","path":null,"label":"NousResearch/DisTrO","type":"external"},{"url":"https://huggingface.co/NousResearch/Hermes-4.3-36B","path":null,"label":"NousResearch/Hermes-4.3-36B","type":"external"},{"url":"https://huggingface.co/NousResearch/NousCoder-14B","path":null,"label":"NousResearch/NousCoder-14B","type":"external"},{"url":"https://huggingface.co/NousResearch/Hermes-4-70B","path":null,"label":"NousResearch/Hermes-4-70B","type":"external"},{"url":"https://huggingface.co/NousResearch/Minos-v1","path":null,"label":"NousResearch/Minos-v1","type":"external"},{"url":"https://huggingface.co/NousResearch/Hermes-4-14B","path":null,"label":"NousResearch/Hermes-4-14B","type":"external"},{"url":"https://huggingface.co/NousResearch/nomos-1","path":null,"label":"NousResearch/nomos-1","type":"external"},{"url":"https://huggingface.co/NousResearch/Hermes-4-405B","path":null,"label":"NousResearch/Hermes-4-405B","type":"external"},{"url":"https://github.com/NousResearch/hermes-agent/releases/tag/v2026.6.19","path":null,"label":"NousResearch/hermes-agent","type":"external"},{"url":"https://github.com/NousResearch/hermes-agent/releases/tag/v2026.6.5","path":null,"label":"NousResearch/hermes-agent","type":"external"},{"url":"https://nousresearch.com/careers","path":null,"label":"nousresearch.com/careers","type":"external"},{"url":"https://nousresearch.com/careers","path":null,"label":"nousresearch.com/careers","type":"external"},{"url":"https://nousresearch.com/careers","path":null,"label":"nousresearch.com/careers","type":"external"},{"url":"https://nousresearch.com/careers","path":null,"label":"nousresearch.com/careers","type":"external"},{"url":"https://nousresearch.com/careers","path":null,"label":"nousresearch.com/careers","type":"external"},{"url":"https://nousresearch.com/careers","path":null,"label":"nousresearch.com/careers","type":"external"},{"url":"https://github.com/NousResearch/hermes-agent/releases/tag/v2026.5.29.2","path":null,"label":"NousResearch/hermes-agent","type":"external"},{"url":"https://github.com/NousResearch/hermes-agent/releases/tag/v2026.5.29","path":null,"label":"NousResearch/hermes-agent","type":"external"},{"url":"https://github.com/NousResearch/hermes-agent/releases/tag/v2026.5.28","path":null,"label":"NousResearch/hermes-agent","type":"external"},{"url":"https://github.com/NousResearch/Automodel","path":null,"label":"NousResearch/Automodel","type":"external"},{"url":"https://github.com/NousResearch/Megatron-LM","path":null,"label":"NousResearch/Megatron-LM","type":"external"},{"url":"https://github.com/NousResearch/Megatron-Bridge","path":null,"label":"NousResearch/Megatron-Bridge","type":"external"},{"url":"https://github.com/NousResearch/atropos","path":null,"label":"NousResearch/atropos","type":"external"},{"url":"https://github.com/NousResearch/hermes-compression-eval","path":null,"label":"NousResearch/hermes-compression-eval","type":"external"},{"url":"https://github.com/NousResearch/hermes-agent/releases/tag/v2026.5.16","path":null,"label":"NousResearch/hermes-agent","type":"external"},{"url":"https://github.com/NousResearch/hermes-agent-self-evolution","path":null,"label":"NousResearch/hermes-agent-self-evolution","type":"external"},{"url":"https://github.com/NousResearch/agent-governance-toolkit","path":null,"label":"NousResearch/agent-governance-toolkit","type":"external"},{"url":"https://huggingface.co/NousResearch/DeepHermes-ToolCalling-Specialist-Atropos","path":null,"label":"NousResearch/DeepHermes-ToolCalling-Specialist-Atropos","type":"external"},{"url":"https://huggingface.co/NousResearch/Llama-3.2-1B","path":null,"label":"NousResearch/Llama-3.2-1B","type":"external"},{"url":"https://github.com/NousResearch/hermes-agent/releases/tag/v2026.5.7","path":null,"label":"NousResearch/hermes-agent","type":"external"},{"url":"https://github.com/NousResearch/kanban-video-pipeline","path":null,"label":"NousResearch/kanban-video-pipeline","type":"external"},{"url":"https://huggingface.co/NousResearch/DeepHermes-Financial-Fundamentals-Prediction-Specialist-Atropos","path":null,"label":"NousResearch/DeepHermes-Financial-Fundamentals-Prediction-Specialist-Atropos","type":"external"},{"url":"https://github.com/NousResearch/hermes-agent/releases/tag/v2026.4.30","path":null,"label":"NousResearch/hermes-agent","type":"external"},{"url":"https://huggingface.co/NousResearch/moe-10b-a1b-8k-wsd-lr3e4-1t","path":null,"label":"NousResearch/moe-10b-a1b-8k-wsd-lr3e4-1t","type":"external"},{"url":"https://github.com/NousResearch/Nemotron","path":null,"label":"NousResearch/Nemotron","type":"external"},{"url":"https://github.com/NousResearch/NemoClaw","path":null,"label":"NousResearch/NemoClaw","type":"external"},{"url":"https://github.com/NousResearch/hermes-agent/releases/tag/v2026.4.23","path":null,"label":"NousResearch/hermes-agent","type":"external"},{"url":"https://github.com/NousResearch/wterm","path":null,"label":"NousResearch/wterm","type":"external"},{"url":"https://github.com/NousResearch/OpenShell","path":null,"label":"NousResearch/OpenShell","type":"external"},{"url":"https://github.com/NousResearch/OpenShell-Community","path":null,"label":"NousResearch/OpenShell-Community","type":"external"},{"url":"https://github.com/NousResearch/hermes-agent/releases/tag/v2026.4.16","path":null,"label":"NousResearch/hermes-agent","type":"external"},{"url":"https://huggingface.co/NousResearch/DeepHermes-AscensionMaze-RLAIF-8b-Atropos","path":null,"label":"NousResearch/DeepHermes-AscensionMaze-RLAIF-8b-Atropos","type":"external"},{"url":"https://github.com/NousResearch/hermes-paperclip-adapter","path":null,"label":"NousResearch/hermes-paperclip-adapter","type":"external"},{"url":"https://github.com/NousResearch/hermes-agent/releases/tag/v2026.4.13","path":null,"label":"NousResearch/hermes-agent","type":"external"},{"url":"https://github.com/NousResearch/ink","path":null,"label":"NousResearch/ink","type":"external"},{"url":"https://github.com/NousResearch/RL","path":null,"label":"NousResearch/RL","type":"external"},{"url":"https://github.com/NousResearch/Gym","path":null,"label":"NousResearch/Gym","type":"external"},{"url":"https://github.com/NousResearch/hermes-agent/commit/2666009ccc00f653a1db4ff249c9b1aabdb34f13","path":null,"label":"NousResearch/hermes-agent","type":"external"},{"url":"http://github.com/NousResearch/hermes-agent/releases/tag/v2026.6.5","path":null,"label":"NousResearch/hermes-agent","type":"external"},{"url":"https://www.marktechpost.com/2026/05/16/nous-research-proposes-lighthouse-attention-a-training-only-selection-based-hierarchical-attention-that-delivers-1-4-1-7x-pretraining-speedup-at-long-context/","path":null,"label":"marktechpost.com/2026","type":"external"},{"url":"https://newsference.com/c/nous-lighthouse-attention-long-context-speedup","path":null,"label":"newsference.com/c","type":"external"},{"url":"https://github.com/nousresearch","path":null,"label":"github.com","type":"external"},{"url":"https://the-decoder.com/nous-research-releases-hermes-desktop-an-open-source-ai-agent-for-every-platform/","path":null,"label":"the-decoder.com/nous-research-releases-hermes-desktop-an-open-source-ai-agent-for-every-platform","type":"external"},{"url":"https://huggingface.co/NousResearch/Hermes-3-Llama-3.1-70B","path":null,"label":"NousResearch/Hermes-3-Llama-3.1-70B","type":"external"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-pro","workflow":"onlylabs-deepagents-analysis-v3","agent":"deepagents"},"evidence":{"total":94,"pages":28,"events":140,"web":6,"signal_desks":{"forks":12,"repos":9,"hiring":6,"talking":0,"releases":33},"data_radar_lanes":null,"data_radar_matches":null}},"signal_counts":{"total":150,"model_released":30,"release":23,"repo_new":39,"repo_forked":46,"post_published":0,"job_opened":12}}