Models

BH P300

Device keys in tt-inference-server: P300.

Community models are tt-model bundles published by their owners and are not in tt-inference-server's release spec. Tenstorrent has not reviewed them. Serve at your own risk.

19 models
Image generationOfficial

FLUX.1-dev

black-forest-labs/FLUX.1-dev

Listed in tt-inference-server on 6 hardware types; best status Complete.

Runs on
BH 4xP150, BH LoudBox, BH P300 +3
Status
Complete
tt launch
None
Image generationOfficial

FLUX.1-schnell

black-forest-labs/FLUX.1-schnell

Listed in tt-inference-server on 6 hardware types; best status Complete.

Runs on
BH 4xP150, BH LoudBox, BH P300 +3
Status
Complete
tt launch
None
LLMOfficial

Llama-3.1-8B

meta-llama/Llama-3.1-8B

Listed in tt-inference-server on 8 hardware types; best status Complete.

Runs on
BH LoudBox, BH P300, BH QuietBox 2 +5
Speed
≥ 85 tok/s≤ 46 ms WH LoudBox/QuietBox
Status
Complete
tt launch
7 clients
8B parameters
LLMOfficial

Llama-3.1-8B-Instruct

meta-llama/Llama-3.1-8B-Instruct

Listed in tt-inference-server on 9 hardware types; best status Complete.

Runs on
BH LoudBox, BH P300, BH QuietBox 2 +6
Speed
≥ 36 tok/s≤ 58 ms WH LoudBox/QuietBox
Status
Complete
tt launch
7 clients
8B parameters
LLM (inferred)Community model

GLM-4.5-Air

zai-org/GLM-4.5-Air

Published as 1 tt-model bundle by raahemnabeel.

Runs on
BH 4xP150, BH P300, BH QuietBox 2 +1
Status
Not in the release spec
tt launch
7 clients
1 tt-model bundle
LLM (inferred)Community model

Kimi-Linear-48B-A3B-Instruct

moonshotai/Kimi-Linear-48B-A3B-Instruct

Published as 1 tt-model bundle by jashansinghTT.

Runs on
BH P300, BH QuietBox 2
Status
Not in the release spec
tt launch
7 clients
1 tt-model bundle
Served with vLLM, task not statedCommunity model

MIRA-SFT-35B-A3B

MIRA-MMDeepResearch/MIRA-SFT-35B-A3B

MIRA-SFT-35B-A3B, a multimodal deep-research agent model: Qwen3.6-35B-A3B (35B-parameter mixture of experts, 3B active per token: 256 experts, top-8 plus a shared expert; 40 layers mixing Gated DeltaNet linear attention with gated full…

Runs on
BH 4xP150, BH P300, BH QuietBox 2 +3
Status
Not in the release spec
tt launch
7 clients
1 tt-model bundle
Served with vLLM, task not statedCommunity model

MiroThinker-1.7-mini

miromind-ai/MiroThinker-1.7-mini

MiroThinker-1.7-mini, a deep-research agent model: Qwen3-30B-A3B (30B-parameter mixture of experts, 3B active per token: 128 experts, top-8, 48 layers) trained for long-horizon search, browsing and tool use (up to 300 tool calls per task).

Runs on
BH 4xP150, BH P300, BH QuietBox 2 +2
Status
Not in the release spec
tt launch
7 clients
1 tt-model bundle
LLMCommunity model

Olmo-3.1-32B-Instruct

allenai/Olmo-3.1-32B-Instruct

Olmo 3.1 32B Instruct, a fully open 32B dense decoder with hybrid sliding-window / full attention and 65k context, served with vLLM.

Runs on
BH P300, BH QuietBox 2, p150
Speed
11 tok/s177 ms 1 chip
Status
Not in the release spec
tt launch
7 clients
1 tt-model bundle
LLM (inferred)Community model

Ornith-1.5-35B-A3B

ornith-ai/Ornith-1.5-35B-A3B

Ornith-1.5-35B-A3B, a 35B-parameter mixture-of-experts model with 3B parameters active per token (256 experts, top-8 plus a shared expert; 40 layers mixing Gated DeltaNet linear attention with gated full attention) and a vision encoder,…

Runs on
BH 4xP150, BH P300, BH QuietBox 2 +3
Speed
220 tok/s118 ms 4 chips
Status
Not in the release spec
tt launch
7 clients
1 tt-model bundle
Served with vLLM, task not statedCommunity model

QUEST-35B-RL

osunlp/QUEST-35B-RL

QUEST-35B-RL, a deep-research agent model: Qwen3.5-35B-A3B (35B-parameter mixture of experts, 3B active per token: 256 experts, top-8 plus a shared expert; 40 layers mixing Gated DeltaNet linear attention with gated full attention; vision…

Runs on
BH 4xP150, BH P300, BH QuietBox 2 +3
Status
Not in the release spec
tt launch
7 clients
1 tt-model bundle
LLM (inferred)Community model

Qwen3-Coder-30B-A3B-Instruct

Qwen/Qwen3-Coder-30B-A3B-Instruct

Qwen3-Coder-30B-A3B-Instruct (30B total / 3B active MoE: 128 experts, top-8, 48 layers) for agentic coding with tool calling on Blackhole through vLLM, with a fused MoE decode: the v1 bundle (raahemnabeel/qwen3-coder-30b-a3b-v1) plus the…

Runs on
BH 4xP150, BH P300, BH QuietBox 2 +2
Speed
96 tok/s145 ms 4 chips
Status
Not in the release spec
tt launch
7 clients
1 tt-model bundle
LLM (inferred)Community model

Qwen3-Coder-Next

Qwen/Qwen3-Coder-Next

Qwen3-Coder-Next (80B total / 3B active MoE: 512 experts, top-10 + shared expert, 48 layers of Gated DeltaNet + gated attention) on Blackhole through vLLM and the Tenstorrent plugin, with a fused decode step: the v1 bundle…

Runs on
BH 4xP150, BH P300, BH QuietBox 2 +1
Speed
179 tok/s173 ms 4 chips
Status
Not in the release spec
tt launch
7 clients
1 tt-model bundle
Text rankingCommunity model

Qwen3-Reranker-4B

Qwen/Qwen3-Reranker-4B

Qwen3-Reranker-4B, a text reranker for 100+ languages and code: it scores how well a document answers a query by reading logit("yes") - logit("no") at the end of the model card's prompt, served as a Cohere / Jina / vLLM-style /v1/rerank…

Runs on
BH 4xP150, BH P300, BH QuietBox 2 +2
Status
Not in the release spec
tt launch
None
1 tt-model bundle
EmbeddingCommunity model

Qwen3-VL-Embedding-8B

Qwen/Qwen3-VL-Embedding-8B

Qwen3-VL-Embedding-8B, a multimodal embedding model: text, images, screenshots, video, or a mix, map to one L2-normalised 4096-dimensional vector in a shared space (Matryoshka dimensions down to 64), served as an OpenAI-style…

Runs on
BH 4xP150, BH P300, BH QuietBox 2 +2
Status
Not in the release spec
tt launch
None
1 tt-model bundle
Text rankingCommunity model

Qwen3-VL-Reranker-8B

Qwen/Qwen3-VL-Reranker-8B

Qwen3-VL-Reranker-8B, a multimodal reranker: it scores how well a document answers a query, where either side can be text, images, screenshots, video or a mix.

Runs on
BH 4xP150, BH P300, BH QuietBox 2 +2
Status
Not in the release spec
tt launch
None
1 tt-model bundle
LLM (inferred)Community model

Qwen3.6-35B-A3B

Qwen/Qwen3.6-35B-A3B

Published as 2 tt-model bundles by Adartras, rothgar.

Runs on
BH 4xP150, BH P300, BH QuietBox 2 +2
Status
Not in the release spec
tt launch
7 clients
2 tt-model bundles
Text-to-speechCommunity model

s2-pro

fishaudio/s2-pro

Fish Audio S2 Pro, a 4.5B dual-autoregressive text-to-speech model (Qwen3-based semantic decoder, 400M codebook decoder, 44.1 kHz DAC codec) on Tenstorrent Blackhole, with a fish-speech-compatible /v1/tts API (streaming WAV, voice cloning…

Runs on
BH P300, BH QuietBox 2, p100 +1
Status
Not in the release spec
tt launch
None
2 tt-model bundles
Served with vLLM, task not statedCommunity model

Tongyi-DeepResearch-30B-A3B

Alibaba-NLP/Tongyi-DeepResearch-30B-A3B

Tongyi DeepResearch, an agent model for long-horizon, deep information seeking: a 30B-parameter mixture of experts with 3B active per token (Qwen3-30B-A3B architecture: 128 experts, top-8, 48 layers), trained with agentic continual…

Runs on
BH 4xP150, BH P300, BH QuietBox 2 +2
Status
Not in the release spec
tt launch
7 clients
1 tt-model bundle