For when 7B is no longer enough and 70B is not needed yet. Noticeably sharper at code and maths, invents fewer facts, and still runs on a single card — no sharding, no multi-GPU dance.
| MMLU | 83.3 | |
|---|---|---|
| GSM8K | 95.9 | |
| HumanEval | 88.4 |
Which GPU does Qwen2.5 32B need?
The entry configuration is 1x RTX5090, which delivers around 38 tokens per second. For more parallel requests and headroom on speed, go with 1x H100.
How much does Qwen2.5 32B cost?
An hour of Qwen2.5 32B runs from $0.65 to $3.33 depending on the GPU configuration. Billing is hourly: you pay for server time, not for tokens.
How do I call Qwen2.5 32B from my code?
Once deployed you get an OpenAI-compatible endpoint: point any OpenAI client at the base_url you receive and pass qwen2.5:32b as the model name. Code written against the OpenAI API needs no changes.
Does the data stay private?
Yes. Qwen2.5 32B runs on a dedicated GPU server inside a private loop: requests and responses never reach the model vendor and are not used for training.
What context window and license does Qwen2.5 32B have?
Qwen2.5 32B holds 32K tokens of context per request and ships under the Apache 2.0 license, so you can use it commercially on your own hosting terms.