Model Pools

Nodes grouped by model tier with shared pricing. Pick a tier, and the network handles routing.

How Pools Work: Operators join a pool based on the model tier they serve. Revenue is split proportionally by contribution (tokens served, uptime, speed). Platform fee: 15%.

Lite Pool

Quick tasks, autocomplete, lightweight inference

$0.001

per 1K tokens

Members

850 t/s

Throughput

65ms

Latency

99.5%

Uptime

1.2M requests served

Standard Pool

General chat, code completion, reasoning tasks

$0.002

per 1K tokens

Members

520 t/s

Throughput

120ms

Latency

99.2%

Uptime

3.4M requests served

Pro Pool

14B

Code generation, analysis, complex tasks

$0.005

per 1K tokens

Members

280 t/s

Throughput

180ms

Latency

98.8%

Uptime

890K requests served

Premium Pool

32B

Complex reasoning, long-form content, demanding analysis

$0.012

per 1K tokens

Members

90 t/s

Throughput

350ms

Latency

98.5%

Uptime

245K requests served