usepod-agent 0.3.3

Use Pod marketplace provider agent — connects local inference backends (vLLM, llama.cpp, LM Studio, Ollama) to the Use Pod coordinator and earns USDC for served jobs.
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36

[operator]
display_name  = "BlueBoxRig-01"
wallet        = "11111111111111111111111111111111"
contact_email = "ops@example.com"

[coordinator]
url             = "wss://api.usepod.ai/provider/connect"
enrollment_code = "POD-ENROLL-7K9X4LM2"

[identity]
key_path = "~/.usepod-agent/identity.key"

[[backends]]
kind = "vllm"
url  = "http://localhost:8000"

[[backends]]
kind = "openrouter"
api_key_env = "OPENROUTER_API_KEY"
markup = 0.05

[pricing]
default_input_per_1m  = 500_000
default_output_per_1m = 750_000

[pricing.models]
"meta-llama/Llama-3.3-70B-Instruct"  = { input_per_1m = 300_000, output_per_1m = 450_000 }
"mistralai/Mistral-Large"             = { input_per_1m = 800_000, output_per_1m = 1_200_000 }

[limits]
max_concurrent        = 8
max_tokens_per_minute = 200_000

[observability]
prometheus_addr = "127.0.0.1:9090"
log_level       = "info"