~/model/huggingfaceQwen/Qwen2.5-Coder-32B-Instruct
Alibaba provider

Qwen 2.5 Coder 32B

Code-specialized 32B. Near-GPT-4 code completion and repair at local-inference cost.

Parameters
32.5B
Family
Alibaba
Context
128K tokens
FP16 weights
65GB
// where you can run it
qwen2.5-coder:32bLM StudiovLLMMLXoMLX
// hugging face stats (cached daily)
764K downloads · 2K likes · license: apache-2.0 · updated 1 year ago

What you need to run this.

$ ./vrambudget --model qwen-2-5-coder-32b --by quant
FP1665 GB
H100 80GB80 GB · fitsM3 Max 9696 GB · fitsRTX Pro 600096 GB · fitsM4 Max 128128 GB · fits
Q8_035 GB
RTX A600048 GB · fitsRTX 6000 Ada48 GB · fitsL40S48 GB · fitsRadeon Pro W790048 GB · fits
Q6_K27 GB
RTX A600048 GB · fitsRTX 6000 Ada48 GB · fitsL40S48 GB · fitsRadeon Pro W790048 GB · fits
Q5_K_M22 GB
RTX 509032 GB · fitsRTX A600048 GB · fitsRTX 6000 Ada48 GB · fitsL40S48 GB · fits
Q4_K_M18 GB
RTX 309024 GB · fitsRTX 3090 Ti24 GB · fitsRTX 409024 GB · fitsRX 7900 XTX24 GB · fits

// budgets shown at ctx 8K, concurrency 1, 15% safety headroom. Tune in the calculator →

Alternatives at this size.

$ grep --params similar catalog.json
$ similarQwen 2.5 32B32.5BAlibaba$ similarYi 34B34B01.AI$ similarQwen3 30B A3B30.5BAlibaba$ similarQwen 3.6 35B A3B35BAlibaba

Discussion.

$ gh discussion list

// sign in with github to leave a comment. threads live in the repo's discussions tab.