GLM-5.2 is live. #1 throughput on OpenRouter, pay-per-token on FriendliAI. Try it today ➜
FINAL-Bench
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
FINAL-Bench
Model Tree
Qwen/Qwen3.5-4B
FINAL-Bench/Darwin-4B-David
Input Modalities
Output Modalities
Supported Functionality
Darwin-4B-Genesis is the 3rd generation Darwin model and the world's first model to successfully crossbreed FFN layers across different architectures — Transformer (Gemma4) and Mamba (Qwen3.5 GatedDeltaNet) — using evolutionary optimization.
The father's Attention layers (Gemma4 Transformer) are preserved at 100%, while the mother's FFN knowledge (Qwen3.5 Mamba) is transplanted at layer-specific optimal ratios discovered automatically by CMA-ES across 42 dimensions.
The result: the child outperforms both parents on every benchmark — a phenomenon known as Hybrid Vigor.
Existing hybrid models (Jamba, Nemotron-H, Granite 4.0) are all designed and trained from scratch. Darwin-4B-Genesis takes two already-trained models from different architecture families and breeds them evolutionarily — with zero additional training.
| Benchmark | David (Father) | Qwen3.5-4B (Mother) | Genesis (Child) |
|---|---|---|---|
| CLIcK | 90% | ~50% (est.) | 92% ✅ |
| MuSR | 65% | ~55% (est.) | 70% ✅ |
The child surpasses both parents. This is the first demonstration of Hybrid Vigor in AI model breeding.
| Benchmark | Genesis | David (Gen2) | K-AI #1 (27B) |
|---|---|---|---|
| CLIcK (Korean culture) | 92% | 90% | 0.794 |
| MuSR (multi-step reasoning) | 70% | 65% | 0.604 |
| GPQA (deep reasoning) | ~60% | ~60% | — |
markdown
Father: Darwin-4B-David (Gemma4 Transformer, hidden=2560, 42 layers)Mother: Qwen/Qwen3.5-4B (GatedDeltaNet/Mamba, hidden=2560, 32 layers)Key insight: hidden_size matches (2560) → direct FFN replacement possibleMethod: Attention 100% from Father, FFN blended at per-layer optimal ratiosOptimizer: CMA-ES (Covariance Matrix Adaptation Evolution Strategy)Genome: 42 dimensions (one ratio per layer)Fitness: CLIcK 60% + MuSR 40% composite scoreFrozen layers: L15, L16, L22, L23, L24, L25 (Korean language preservation)
markdown
L00: 0.206 ██████████░ 21% QwenL07: 0.000 ░░░░░░░░░░░ Auto-protected by CMA-ESL15: 0.000 ░░░░░░░░░░░ Frozen (Korean)L22: 0.000 ░░░░░░░░░░░ Frozen (Korean)L29: 0.291 ██████████████░ 29% Qwen (maximum)L31: 0.244 ████████████░ 24% QwenL32: 0.273 █████████████░ 27% Qwen
Key finding: CMA-ES applied the most aggressive Qwen blending to the final layers (L29-32), which govern output quality.
python
from transformers import AutoTokenizer, AutoModelForCausalLMtokenizer = AutoTokenizer.from_pretrained("FINAL-Bench/Darwin-4B-Genesis",trust_remote_code=True,)model = AutoModelForCausalLM.from_pretrained("FINAL-Bench/Darwin-4B-Genesis",dtype="bfloat16",device_map="auto",trust_remote_code=True,)messages = [{"role": "user", "content": "Explain how hybrid vigor works in genetics."}]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=1024, do_sample=False)print(tokenizer.decode(outputs[0][inputs['input_ids'].shape[-1]:], skip_special_tokens=True))
markdown
google/gemma-4-E4B-it × TeichAI/Claude-Opus-Distill-E4B→ Darwin-4B-Opus (Gen 1, DARE-TIES merge)Darwin-4B-Opus × DavidAU/DECKARD-Expresso-Universe→ Darwin-4B-David (Gen 2, MRI-guided merge, CLIcK 90%)Darwin-4B-David × Qwen/Qwen3.5-4B→ Darwin-4B-Genesis (Gen 3, Cross-Arch FFN Breeding, CLIcK 92%) ★
bibtex
@misc{vidraft_darwin_4b_genesis,title = {Darwin-4B-Genesis: World's First Cross-Architecture FFN Breeding},author = {VIDRAFT},year = {2026},publisher = {Hugging Face},howpublished = {\url{https://huggingface.co/FINAL-Bench/Darwin-4B-Genesis}}}@article{kim2026darwin,title={Darwin Family: MRI-Trust-Weighted Evolutionary Merging for Training-Free Scaling of Language-Model Reasoning},author={Kim, Taebong and Hong, Youngsik and Kim, Minsik and Choi, Sunyoung and Jang, Jaewon and Shin, Junghoon and Kim, Minseo},journal={arXiv preprint arXiv:2605.14386},year={2026}}