Serve on Badgr
Serve Llama 3.1 8B Instruct
Meta's flagship 8B instruction model — strong general-purpose baseline.
Recommended GPU
RTX 4090 24GB
Available regions
United States, Europe
Estimated cost range
$0.21 – $0.23/hr
Startup time
2–5 min
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="llama-8b",
messages=[{"role": "user", "content": "Hello!"}],
)
print(response.choices[0].message.content)Run command
badgr serve llama-8b \
--gpu RTX 4090 \
--max-cost 10Context length: 128K tokens · Model ID: meta-llama/Llama-3.1-8B-Instruct
Recommended Badgr routes
RTX 4090 24GB · United States
Availablefrom $0.21/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 $0.21/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 $0.21/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 $0.21/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 $0.23/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 $0.23/hr
Best for
ComfyUI, Batch Inference, Qwen 2.5 7B Instruct
Startup: 2–5 min
Reliability: Standard
Last checked: just now