Qwen3-32B

Listed in tt-inference-server on 5 hardware types; best status Complete. 3 tt-model bundles also published.

In tt-inference-server

Hardware	Engine and implementation	Status	Context	Users	Build	Support page
WH Galaxy `GALAXY`	vLLM qwen3-32b-galaxy	Complete	128K	32	Code: https://github.com/tenstorrent/tt-metal/tree/bc4c4df/models/demos/llama3_70b_galaxy tt-metal `bc4c4df` vLLM `be7d805` Image `ghcr.io/tenstorrent/tt-inference-server/vllm-tt-metal-src-release-ubuntu-22.04-amd64:0.21.0-bc4c4df-be7d805`	Open
BH Galaxy `BLACKHOLE_GALAXY`	vLLM qwen3-32b-galaxy	Functional	128K	32	Code: https://github.com/tenstorrent/tt-metal/tree/978f93b/models/demos/llama3_70b_galaxy tt-metal `978f93b` vLLM `a723030` Image `ghcr.io/tenstorrent/tt-inference-server/vllm-tt-metal-src-release-ubuntu-22.04-amd64:0.25.0-978f93b-a723030`	Open
BH LoudBox `P150X8`	vLLM tt-transformers	Functional	128K	32	Code: https://github.com/tenstorrent/tt-metal/tree/555f240/models/tt_transformers tt-metal `555f240` vLLM `22be241` Image `ghcr.io/tenstorrent/tt-inference-server/vllm-tt-metal-src-release-ubuntu-22.04-amd64:0.10.0-555f240-22be241`	Open
BH QuietBox 2 `P300X2`	vLLM tt-transformers	Functional	128K	32	Code: https://github.com/tenstorrent/tt-metal/tree/8c48a10/models/tt_transformers tt-metal `8c48a10` vLLM `f52987a` Image `ghcr.io/tenstorrent/tt-inference-server/vllm-tt-metal-src-release-ubuntu-22.04-amd64:0.17.0-8c48a10-f52987a`	Open
WH Galaxy `GALAXY`	vLLM tt-transformers	Functional	128K	128	Code: https://github.com/tenstorrent/tt-metal/tree/e95ffa5/models/tt_transformers tt-metal `e95ffa5` vLLM `48eba14` Image `ghcr.io/tenstorrent/tt-inference-server/vllm-tt-metal-src-release-ubuntu-22.04-amd64:0.4.0-e95ffa5-48eba14`	Open
WH Galaxy `GALAXY_T3K`	vLLM tt-transformers	Functional	128K	32	Code: https://github.com/tenstorrent/tt-metal/tree/e95ffa5/models/tt_transformers tt-metal `e95ffa5` vLLM `48eba14` Image `ghcr.io/tenstorrent/tt-inference-server/vllm-tt-metal-src-release-ubuntu-22.04-amd64:0.4.0-e95ffa5-48eba14`	-
WH LoudBox/QuietBox `T3K`	vLLM tt-transformers	Functional	128K	32	Code: https://github.com/tenstorrent/tt-metal/tree/e95ffa5/models/tt_transformers tt-metal `e95ffa5` vLLM `48eba14` Image `ghcr.io/tenstorrent/tt-inference-server/vllm-tt-metal-src-release-ubuntu-22.04-amd64:0.4.0-e95ffa5-48eba14`	Open

Status, context and users are copied from tt-inference-server's release spec (commit 2db98e38c, release 0.25.0). Context is the longest prompt plus reply the entry allows. Users is its highest concurrency. Each row carries its own code link, tt-metal and vLLM commits and image, because they differ between hardware. Nothing on this page is measured.

tt-model bundles

Bundle	Hardware	Served with	Context	Pushed
changh95/qwen3-32b-bh-galaxy	BH Galaxy	Not stated	128K	2026-10-02
changh95/qwen3-32b-wh-galaxy	n150x32	vllm-plugin	128K	2026-10-02
mando2222/qwen3-32b-blackhole-v51	BH 4xP150	vllm-plugin	32K	2026-09-09

From tt-model search --catalog and each bundle's manifest (tt-model info). A bundle lists every hardware profile its manifest defines. Each bundle is published by its owner, and Tenstorrent has not reviewed it. Serve at your own risk.

Details

Kind	LLM (from tt-inference-server model type)
Parameters	32B
Family	Qwen (from the start of the name)
Implementation	qwen3-32b-galaxy, tt-transformers
Disk needed	84 GB
RAM needed	80 GB
Tool-call parser	`hermes`
Reasoning parser	`qwen3`
Listed by	tt-inference-server

Install and run

Download the weights

tt model pull Qwen3-32B

Serve it on this machine

tt serve Qwen3-32B

To serve one of its tt-model bundles instead, use the Copy serve button beside it.

tt-studio opens this as the Chat page. It picks the page from the model type.

Connect a client with tt launch

Serve the model first, then run one of these. Only clients that work with this model are listed.