Instructions to use moondream/moondream-next with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use moondream/moondream-next with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="moondream/moondream-next", trust_remote_code=True)

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained("moondream/moondream-next", trust_remote_code=True, device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use moondream/moondream-next with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "moondream/moondream-next"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "moondream/moondream-next",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/moondream/moondream-next

SGLang

How to use moondream/moondream-next with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "moondream/moondream-next" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "moondream/moondream-next",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "moondream/moondream-next" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "moondream/moondream-next",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use moondream/moondream-next with Docker Model Runner:
```
docker model run hf.co/moondream/moondream-next
```

err805 commited on Jun 19

Commit

9ca19fc

verified ·

1 Parent(s): 71571da

Upload merged OPD RL weights

Browse files

Files changed (3) hide show

manifest.json +36 -0
model.pt +3 -0
model.pt.sha256 +1 -0

manifest.json ADDED Viewed

	@@ -0,0 +1,36 @@

+{
+  "base_model": "/root/.cache/huggingface/hub/models--moondream--moondream3-preview/snapshots/5112966d1a723413b1c9a1e8bea272b72e647b35/model_fp8.pt",
+  "lora_ckpt": "/workspace/evals/merged-opd-rl/source/multi-rl-from-opd-dual-r512-p64-step000004002-ckpt.pt",
+  "source_run": "multi-rl-from-opd-dual-r512-p64-4k-20260617-235617",
+  "source_step": 4002,
+  "output": "/workspace/evals/merged-opd-rl/output/moondream3-preview__multi-rl-from-opd-dual-r512-p64-4k-20260617-235617__step-000004002/model.pt",
+  "dense_layers": [
+    0,
+    1,
+    2,
+    3
+  ],
+  "moe_layers": [
+    4,
+    5,
+    6,
+    7,
+    8,
+    9,
+    10,
+    11,
+    12,
+    13,
+    14,
+    15,
+    16,
+    17,
+    18,
+    19,
+    20,
+    21,
+    22,
+    23
+  ],
+  "format": "md3_flat_bf16_moe_merged"
+}

model.pt ADDED Viewed

	@@ -0,0 +1,3 @@

+version https://git-lfs.github.com/spec/v1
+oid sha256:55149f2a0999a26b84fe4af63c24782548fdcf86610e21b768f0b381a28bb373
+size 18537525163

model.pt.sha256 ADDED Viewed

	@@ -0,0 +1 @@


1	+ 55149f2a0999a26b84fe4af63c24782548fdcf86610e21b768f0b381a28bb373 /workspace/evals/merged-opd-rl/output/moondream3-preview__multi-rl-from-opd-dual-r512-p64-4k-20260617-235617__step-000004002/model.pt