GGUF models by cstr
462 GGUF models quantized by cstr on Hugging Face. Sizes run from 1M to 10.7B parameters. The longest trained context is 128K tokens.
About cstr on Hugging Face
cstr has published 462 GGUF models that ggufmodels.com tracks, across families such as Opus, Stt and Qwen. The most downloaded are ark asr 3b, qwen3 tts 0.6b customvoice and qwen3-tts-tokenizer-12hz.
The quants cstr publishes most often are Q4_K (312), Q8_0 (306), F16 (252) and IQ4_XS (65). Their Q4_K_M files range from 0.8 to 5.3 GB. Parameter counts run from 1M to 10.7B, and the longest trained context window is 128K tokens.
Frequently asked questions
What GGUF models has cstr published?
462 models are listed here, mainly from the Opus, Stt and Qwen families. The most downloaded is ark asr 3b.
Which quants does cstr publish?
The most common are Q4_K (312 models), Q8_0 (306 models), F16 (252 models) and IQ4_XS (65 models).