NOUS RELEASES

PROJECT NAMETYPEDETAILSRELEASE DATESIZE
1Measuring Thinking Efficiency in Reasoning ModelsRESEARCHMeasuring Thinking Efficiency in Reasoning Models: The Missing Benchmark08/14/25
2Hermes 3 DatasetDATASETComplete dataset used in pretraining of Hermes 3 models07/11/25
3Steering the ShoggothsRESEARCHTaming LLMs with Sequential Monte Carlo06/01/25
4Psyche NetworkTRAININGOpen infrastructure democratizing AI development05/14/25
5AtroposFRAMEWORKLanguage Model RL Environments04/29/25
6DeepHermes-3-Llama-3-8B-PreviewMODELModel with deep reasoning and chat modes02/14/2516.1
7Hermes-3-Llama-3.2-3BMODEL
Model that can run on most modern hardware12/11/246.5
8GODSIMULATORSocial agent11/29/24
9S8NSIMULATORSocial agent11/29/24
10DeMo: Decoupled Momentum OptimizationPAPERTechnical report11/29/24
11Hermes ChatSIMULATORChat Interface11/08/24
12TEESIMULATORAutonomous agent10/29/24
13A Preliminary Report On DisTrOPAPERTechnical report08/26/24
14Original Nous WorldSim PromptPROMPTPrompt engineering03/15/24
15Hermes 3 Technical ReportPAPERTechnical report08/14/24
16Hermes 2 Pro 70BMODELFunction-calling model06/27/2442.5
17Hermes 2 ThetaMODELModel06/20/2442.5
18Character CodexDATASETRoleplaying dataset06/11/240.0112
19Nous WorldSimCODECLI simulator03/23/24
20Hermes-2-Pro-Mistral-7BMODELMaster function-calling model03/11/202414.49
21NousResearch/Genstruct-7BMODELData synthesis model03/07/202415
22Nous Research Official Branding BookletPAPERDesign & story of Nous03/05/20240.042
23NousResearch/Nous-Hermes-2-Mistral-7B-DPOMODEL7B model02/20/202415
24NousResearch/finetuning-subnetSUBNETEvals on-chain01/29/20240.0001
25NousResearch/Nous-Hermes-2-Mixtral-8x7B-DPOMODELMoE model + DPO01/15/202495
26NousResearch/Nous-Hermes-2-Mixtral-8x7B-SFTMODELMoE model - DPO01/15/202495
27NousResearch/Nous-Hermes-2-Mixtral-8x7B-DPO-GGUFMODELQuantized MoE model01/15/202417-50
28NousResearch/Nous-Hermes-2-Yi-34BMODELHermes on Yi12/25/202370
29NousResearch/Nous-Capybara-34BMODELCapybara-series model11/14/202370
30NousResearch/Yarn-Mistral-7b-128kMODELLong-context model11/02/202315
31NousResearch/Obsidian-3B-V0.5MODELMultimodal vision LLM10/30/20235
32NousResearch/ObsidianCODECode for obsidian model10/30/20230.015
33YaRN: Efficient Context Window Extension of Large Language ModelsPAPERArchitecture behind YaRN models08/31/20230.0007
34NousResearch/Nous-Hermes-Llama2-13bMODELHermes on Llama207/21/202325

ARTIFICIAL INTELLIGENCE MADE HUMAN

NODES

THE AI ACCELERATOR COMPANY

NODES