Qwen
30 models16 ranked conversions427.8K downloads
Best conversion per model
One row per base model. The full list of attempts is on each model page.
| # | Model | Author | Params | KL | Refusals | Score | Evidence | Downloads |
|---|---|---|---|---|---|---|---|---|
| 1 | Qwenpus3.5-9B-exp sulpikar2 | sulpikar2 | 9B | 0.0031 | 3/1003% | 98.3 | BestParetoSAFETENSORS | 169 |
| 2 | DeepSeek-R1-Distill-Qwen-14B deepseek-ai | blackbook-lm | 14B | 0.0079 | 3/1003% | 98.1 | BestParetoGGUFSAFETENSORS | 11.8K |
| 3 | mythos-qwen-1.5b-final expper | richardyoung | 1.5B | 0.023 | 2/1002% | 97.9 | BestParetoMLXSAFETENSORS | 260 |
| 4 | DeepSeek-R1-Distill-Qwen-7B deepseek-ai | blackbook-lm | 7B | 0.0071 | 5/1005% | 97.2 | BestParetoGGUFSAFETENSORS | 13.2K |
| 5 | DeepSeek-R1-Distill-Qwen-32B deepseek-ai | blackbook-lm | 32B | 0.029 | 3/1003% | 97.1 | BestParetoGGUFSAFETENSORS | 19.7K |
| 6 | Qwen-SEA-LION-v4-4B-VL aisingapore | MagicalAlchemist | 4B | 0.039 | 4/1004% | 96.1 | BestParetoGGUFSAFETENSORS | 9.8K |
| 7 | FuseChat-Qwen-2.5-7B-Instruct FuseAI | ChiKoi7 | 7B | 0.070 | 3/1003% | 95.2 | BestParetoGGUFSAFETENSORS | 25.2K |
| 8 | MiMo-V2.6-Distill-Qwen-9B XiaomiMiMo | saidutta69 | 9B | 0.048 | 8/1008% | 93.7 | BestParetoGGUFSAFETENSORS | 8.5K |
| 9 | QwenPaw-Flash-9B agentscope-ai | coder3101 | 9B | 0.0099 | 12/10012% | 93.5 | BestParetoGGUFMLXSAFETENSORS | 18.8K |
| 10 | Qwen-Image-2.1-PE-T2I Qwen | prithivMLmods | 9.4B | 0.033 | 22/10022% | 87.4 | BestParetoGGUFSAFETENSORS | 1.2K |
| 11 | Qwen-3-VL-Ricoh-8B-20260227 ricoh-ai | sasa2000 | 8B | 0.017 | 27/10027% | 85.7 | BestParetoGGUFSAFETENSORS | 246 |
| 12 | qwen-4b-2507-rp-mahou Pranavz | Pranavz | 4B | 0.420 | 5/1005% | 82.7 | BestParetoGGUFLORASAFETENSORS | 7.2K |
| 13 | Qwen-SEA-LION-v4-8B-VL aisingapore | MagicalAlchemist | 8B | 0.387 | 12/10012% | 80.0 | BestParetoGGUFSAFETENSORS | 8.8K |
All models
Qwenpus3.5-9B-exp9BDeepSeek-R1-Distill-Qwen-14B14Bmythos-qwen-1.5b-final1.5BDeepSeek-R1-Distill-Qwen-7B7BDeepSeek-R1-Distill-Qwen-32B32BQwen-SEA-LION-v4-4B-VL4BFuseChat-Qwen-2.5-7B-Instruct7BMiMo-V2.6-Distill-Qwen-9B9BQwenPaw-Flash-9B9BQwen-Image-2.1-PE-T2I9.4BQwen-3-VL-Ricoh-8B-202602278Bqwen-4b-2507-rp-mahou4BQwen-SEA-LION-v4-8B-VL8BDeepSeek-R1-Distill-Qwen-1.5B1.5BDeepSeek-R1-Distill-Qwen-1.5B-abliterated1.5BDeepseek-R1-Distill-Qwen-32b-uncensored32BHeretic-Nemotron-Research-Reasoning-Qwen-1.5B1.5BHeretic-SLM-Uncensored-Max-Qwen3.1Bqwen-2.5-coder-custom-16bit-heretic7.6BQwen-3.8-27B-Heretic27.8BQwen-AgentWorld-35B-A3B35B · 3B activeQwen-AgentWorld-35B-A3B35B · 3B activeQwen-Fixed-Chat-Templates9.6BQwen-Image-2.1—Qwen-Image-2.1-DiT-NVFP4-ComfyUI6.5BQwen-Image-2.1-PE-I2I9.4BQwen-Image-2.1-Turbo—Qwen1.5-1.8B-Chat-heretic1.8BQwen1.5-4B-Chat-heretic4BQwen1.5-4B-heretic4B
Unmeasured models
DeepSeek-R1-Distill-Qwen-1.5B, DeepSeek-R1-Distill-Qwen-1.5B-abliterated, Deepseek-R1-Distill-Qwen-32b-uncensored, Heretic-Nemotron-Research-Reasoning-Qwen-1.5B, Heretic-SLM-Uncensored-Max-Qwen, qwen-2.5-coder-custom-16bit-heretic, Qwen-3.8-27B-Heretic, Qwen-AgentWorld-35B-A3B, Qwen-AgentWorld-35B-A3B, Qwen-Fixed-Chat-Templates, Qwen-Image-2.1, Qwen-Image-2.1-DiT-NVFP4-ComfyUI, Qwen-Image-2.1-PE-I2I, Qwen-Image-2.1-Turbo, Qwen1.5-1.8B-Chat-heretic, Qwen1.5-4B-Chat-heretic, Qwen1.5-4B-heretic — conversions exist but their publishers did not report KL and refusal numbers, so they are not ranked.