deepseek-ai/DeepSeek-R1-Distill-Qwen-32B

Text Generation • Updated 1 day ago • 75.9k • • 443

Note It takes a lot for a model to impress me nowadays, but this is leaps and bounds better than any other model I ever ran. It's definitely useless for RP, though.

maldv/Qwentile2.5-32B-Instruct

Text Generation • Updated 16 days ago • 1.34k • 31

Note Excellent general Qwen 2.5 32B model. Scores very high in standard testing. It does decently at creative writing and RP. It's steerable, obedient, and shoot above its weight class.

crestf411/Q2.5-32B-Slush

Text Generation • Updated Dec 2, 2024 • 204 • 7

Note Very decent creative writing and RP model. It can adopt different tones for different characters (something EVA struggles with).

EVA-UNIT-01/EVA-Qwen2.5-32B-v0.2

Text Generation • Updated Nov 14, 2024 • 2.9k • 46

Note Good Qwen 2.5 RP fine tune, but its tone is a tad monotonous. Not censored. Unreliable at instruction following and summarizing tasks.

TheDrummer/Cydonia-22B-v1.3

Updated Nov 21, 2024 • 779 • 28

Note Fairly good Mistral Small fine-tune. Metharme for RP, Mistral for brains.

crestf411/MS-sunfall-v0.7.0

Text Generation • Updated Oct 29, 2024 • 240 • 11

Note Creative model and good alternative to Cydonia (Based on Mistral Small 22B). Uses the godawful Mistral instruct format, though.

TheDrummer/Star-Command-R-32B-v1

Updated Sep 2, 2024 • 42 • 24

Note Okay but generic at RP. Decent at task execution. Overall it's kinda 'meh' for a 32B model. The base model's function calling and other advanced prompting methods are still working very well.

ArliAI/Mistral-Small-22B-ArliAI-RPMax-v1.1

Updated Oct 12, 2024 • 517 • 44

Note Better sentence variety than base model. Reduced task execution and understanding as per usual with 'creative' mistral fine-tunes.