GLM-5.3 is live. Run Z.ai's latest model on Friendli Model APIs. Try it today ➜
sovthpaw
Available on FriendliAI
Run this model inference on single tenant GPU with unmatched speed and reliability at scale.
Run this model inference with full control and performance in your environment.
Model Details
Model Provider
sovthpaw
Model Tree
nvidia/Cosmos3-Nano
Qwen/Qwen3-8B
Input Modalities
Output Modalities
Supported Functionality
OMNISENTER BASE 16B IS A DARWIN FAMILY EVOLVED MULTIMODAL MODEL — THE FIRST GENERATION OF THE OMNISENTER LINEAGE. PRODUCED BY FUSING THE REASONING CAPABILITIES OF QWEN3-8B INTO THE MULTIMODAL WORLD MODEL COSMOS3-NANO VIA PER-TENSOR MRI-TRUST FUSION.
THE MODEL PRESERVES ALL OF COSMOS3-NANO'S MULTIMODAL MODALITIES — VISION, AUDIO, VIDEO UNDERSTANDING AND GENERATION — WHILE BLENDING IN QWEN3-8B'S TEXT REASONING STRENGTHS.
| PARENT | ARCHITECTURE | PARAMETERS | ROLE |
|---|---|---|---|
| NVIDIA/COSMOS3-NANO | COSMOS3FORCONDITIONALGENERATION | ~16B | MULTIMODAL WORLD MODEL |
| QWEN/QWEN3-8B | QWEN3FORCAUSALLM | 8B | DENSE TEXT REASONING |
text
METHOD .................. DARWIN FAMILY MRI-TRUST FUSIONGENOME DENSITY (ρ_b) .... 0.5MRI-TRUST COEFF (τ) ..... 0.4TEXT TENSORS MERGED ..... 398COSMOS EXTRAS PRESERVED . 399 (CROSS-ATTN, MOE TWINS, MODALITY)TOTAL OUTPUT TENSORS .... 798SHAPE MATCH RATE ........ 398/398 (100%)MERGE TIME .............. 195SMODEL SIZE .............. 29GB (BFLOAT16, 7 SHARDS)text
OMNISENTER BASE 16B├── TEXT BACKBONE (DARWIN-MERGED QWEN3)│ ├── 36 TRANSFORMER LAYERS│ ├── SELF-ATTN + MLP + NORMS PER LAYER│ ├── EMBED_TOKENS (151,936 VOCAB)│ └── LM_HEAD├── CROSS-MODAL ATTENTION (FROM COSMOS3-NANO)│ ├── ADD_Q/K/V_PROJ + TO_ADD_OUT PER LAYER│ └── NORM_ADDED_Q/K PER LAYER├── MOE GENERATION TWINS (FROM COSMOS3-NANO)│ └── LAYERS.*.MLP_MOE_GEN.* + LAYERNORMS├── VISION ENCODER├── DIFFUSION TRANSFORMER (VIDEO/IMAGE GEN)├── SOUND TOKENIZER└── VAEtext
TEXT REASONING ......... YES — ENHANCED VIA QWEN3-8B FUSIONVISION ................ YES — PRESERVED FROM COSMOS3-NANOAUDIO ................. YES — PRESERVED FROM COSMOS3-NANOVIDEO UNDERSTANDING ... YES — PRESERVED FROM COSMOS3-NANOVIDEO GENERATION ...... YES — PRESERVED FROM COSMOS3-NANOTOOL CALLING .......... BASE CAPABILITY — IMPROVEMENT VIA SFT (PLANNED)AGENTIC BEHAVIOR ...... BASE CAPABILITY — IMPROVEMENT VIA SFT (PLANNED)MUSIC GENERATION ...... NOT YET — ACESTEP INTEGRATION PLANNED (LINE 2)python
from transformers import AutoModelForCausalLM, AutoTokenizerimport torch model = AutoModelForCausalLM.from_pretrained( "sovthpaw/OmniSenter-Base-16B", torch_dtype=torch.bfloat16, device_map="auto",)tokenizer = AutoTokenizer.from_pretrained("sovthpaw/OmniSenter-Base-16B") messages = [{"role": "user", "content": "Hello, what can you do?"}]text = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)inputs = tokenizer([text], return_tensors="pt").to(model.device)outputs = model.generate(**inputs, max_new_tokens=512)print(tokenizer.decode(outputs[0], skip_special_tokens=True))text
HERMES REASONING TOOL USE ............. 5,000 CONVERSATIONSAURETH SFT CURRICULUM ................. 5,000 CONVERSATIONSHERMES AGENT TRACES ................... 3,679 CONVERSATIONSHERMES FUNCTION CALLING + THINKING .... 3,570 CONVERSATIONSHERMES FUNCTION CALLING V1 ............ 1,893 CONVERSATIONS─────────────────────────────────────────────────────────────TOTAL ................................ 34,142 CONVERSATIONSADDITIONAL: NEMOTRON, ATRPOPS, NOUS RESEARCH DATASETS| FORMAT | VRAM | NOTES |
|---|---|---|
| BFLOAT16 (SAFETENSORS) | ~32GB | FULL PRECISION, A100/2×3090 |
| 4-BIT QUANTIZED (QLORA) | ~8GB | FOR FINE-TUNING |
| Q4_K_M GGUF | ~10GB | INFERENCE ON SINGLE 3090 |
text
COSMOS3-NANO ──┐ ├── DARWIN MERGE ──► OMNISENTER BASE 16B (GEN-0)QWEN3-8B ─────┘ │ ├──► GEN-1 (EVOLVED, CMA-ES) ├──► GEN-2 (EVOLVED + SFT) └──► ... CONTINUOUS EVOLUTIONbibtex
@article{darwin2026family, title={Darwin Family: Training-Free Evolutionary Model Merging}, author={Darwin Team}, journal={arXiv preprint arXiv:2605.14386}, year={2026}}TOWARDS SELF-IMPROVEMENT
NOUS RESEARCH
METHOD .................. DARWIN FAMILY MRI-TRUST FUSIONGENOME DENSITY (ρ_b) .... 0.5MRI-TRUST COEFF (τ) ..... 0.4TEXT TENSORS MERGED ..... 398COSMOS EXTRAS PRESERVED . 399 (CROSS-ATTN, MOE TWINS, MODALITY)TOTAL OUTPUT TENSORS .... 798SHAPE MATCH RATE ........ 398/398 (100%)MERGE TIME .............. 195SMODEL SIZE .............. 29GB (BFLOAT16, 7 SHARDS)OMNISENTER BASE 16B├── TEXT BACKBONE (DARWIN-MERGED QWEN3)│ ├── 36 TRANSFORMER LAYERS│ ├── SELF-ATTN + MLP + NORMS PER LAYER│ ├── EMBED_TOKENS (151,936 VOCAB)│ └── LM_HEAD├── CROSS-MODAL ATTENTION (FROM COSMOS3-NANO)│ ├── ADD_Q/K/V_PROJ + TO_ADD_OUT PER LAYER│ └── NORM_ADDED_Q/K PER LAYER├── MOE GENERATION TWINS (FROM COSMOS3-NANO)│ └── LAYERS.*.MLP_MOE_GEN.* + LAYERNORMS├── VISION ENCODER├── DIFFUSION TRANSFORMER (VIDEO/IMAGE GEN)├── SOUND TOKENIZER└── VAETEXT REASONING ......... YES — ENHANCED VIA QWEN3-8B FUSIONVISION ................ YES — PRESERVED FROM COSMOS3-NANOAUDIO ................. YES — PRESERVED FROM COSMOS3-NANOVIDEO UNDERSTANDING ... YES — PRESERVED FROM COSMOS3-NANOVIDEO GENERATION ...... YES — PRESERVED FROM COSMOS3-NANOTOOL CALLING .......... BASE CAPABILITY — IMPROVEMENT VIA SFT (PLANNED)AGENTIC BEHAVIOR ...... BASE CAPABILITY — IMPROVEMENT VIA SFT (PLANNED)MUSIC GENERATION ...... NOT YET — ACESTEP INTEGRATION PLANNED (LINE 2)from transformers import AutoModelForCausalLM, AutoTokenizerimport torch model = AutoModelForCausalLM.from_pretrained( "sovthpaw/OmniSenter-Base-16B", torch_dtype=torch.bfloat16, device_map="auto",)tokenizer = AutoTokenizer.from_pretrained("sovthpaw/OmniSenter-Base-16B") messages = [{"role": "user", "content": "Hello, what can you do?"}]text = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)inputs = tokenizer([text], return_tensors="pt").to(model.device)outputs = model.generate(**inputs, max_new_tokens=512)print(tokenizer.decode(outputs[0], skip_special_tokens=True))HERMES REASONING TOOL USE ............. 5,000 CONVERSATIONSAURETH SFT CURRICULUM ................. 5,000 CONVERSATIONSHERMES AGENT TRACES ................... 3,679 CONVERSATIONSHERMES FUNCTION CALLING + THINKING .... 3,570 CONVERSATIONSHERMES FUNCTION CALLING V1 ............ 1,893 CONVERSATIONS─────────────────────────────────────────────────────────────TOTAL ................................ 34,142 CONVERSATIONSADDITIONAL: NEMOTRON, ATRPOPS, NOUS RESEARCH DATASETSCOSMOS3-NANO ──┐ ├── DARWIN MERGE ──► OMNISENTER BASE 16B (GEN-0)QWEN3-8B ─────┘ │ ├──► GEN-1 (EVOLVED, CMA-ES) ├──► GEN-2 (EVOLVED + SFT) └──► ... CONTINUOUS EVOLUTION@article{darwin2026family, title={Darwin Family: Training-Free Evolutionary Model Merging}, author={Darwin Team}, journal={arXiv preprint arXiv:2605.14386}, year={2026}}