Serve on Badgr
Serve Google: Gemma 3 12B
Google: Gemma 3 12B served as an OpenAI-compatible endpoint on Badgr GPU capacity.
GPU recommendation auto-generated from OpenRouter model data — not hand-verified by Badgr.
Recommended GPU
L40S 48GB
Available regions
United States, Europe, Asia Pacific
Estimated cost range
$0.17 – $0.17/hr
Startup time
2–5 min
Also runs well on A100 40GB, RTX 4090 24GB.
OpenAI-compatible API example
from openai import OpenAI
client = OpenAI(base_url="https://api.aibadgr.com/v1", api_key="YOUR_BADGR_API_KEY")
response = client.chat.completions.create(
model="gemma-3-12b-it",
messages=[{"role": "user", "content": "Hello!"}],
)
print(response.choices[0].message.content)Run command
badgr serve gemma-3-12b-it \
--gpu L40S \
--max-cost 10Context length: 131K tokens · Model ID: google/gemma-3-12b-it
Recommended Badgr routes
RTX 4090 24GB · United States
Availablefrom $0.17/hr
Best for
ComfyUI, Batch Inference, Qwen 2.5 7B Instruct
Startup: 2–5 min
Reliability: Standard
Last checked: just now
RTX 4090 24GB · United States
Availablefrom $0.17/hr
Best for
ComfyUI, Batch Inference, Qwen 2.5 7B Instruct
Startup: 2–5 min
Reliability: Standard
Last checked: just now
RTX 4090 24GB · United States
Availablefrom $0.17/hr
Best for
ComfyUI, Batch Inference, Qwen 2.5 7B Instruct
Startup: 2–5 min
Reliability: Standard
Last checked: just now
RTX 4090 24GB · Europe
Availablefrom $0.17/hr
Best for
ComfyUI, Batch Inference, Qwen 2.5 7B Instruct
Startup: 2–5 min
Reliability: Standard
Last checked: just now
RTX 4090 24GB · Europe
Availablefrom $0.17/hr
Best for
ComfyUI, Batch Inference, Qwen 2.5 7B Instruct
Startup: 2–5 min
Reliability: Standard
Last checked: just now
RTX 4090 24GB · Europe
Availablefrom $0.17/hr
Best for
ComfyUI, Batch Inference, Qwen 2.5 7B Instruct
Startup: 2–5 min
Reliability: Standard
Last checked: just now