ROI calculator

Model the GPU economics behind coding agents.

Estimate capacity, concurrency, and monthly savings when you move high-volume coding-agent inference from API spend to managed GPUs running the Subconscious runtime.

Share by URL

Every calculator assumption lives in the link. Copy it to share a scenario or bookmark it to reopen the same model later.

Calculator workspace

Tune assumptions and compare scenarios.

Foreground tokens / day

3.3B

Engineer-facing HITL demand

Background tokens / day

833.3M

Low-priority fill work

100%

Agent trace characteristics

FieldValue
Turns per trace
Tokens added / turn
Tool overhead / turn
sec

Unique tokens / trace

695.3K

Total tokens / trace

45.9M

Average time / trace

9.6 min

GPU time: 510.2 sec

Traces / day

4.8K

Avg active traces: 32.1

Peak active traces

76.9

Peak traces/hour: 479.4

Burst metrics

FieldValue
Peak traces / hr multiplier
x
Burst concurrent sessions / engineer
x

Average burst candidate

115.4

1.5x × 76.9 peak active traces

Engineer floor

100

1.0x × 100 engineers

Peak burst active traces

115.4

max(115.4, 100)

Replica planning

Recommended replicas

4

16 GPUs

Override replicas

Uses the recommendation until you enter an override.

Concurrency capacity

128

Total TPS capacity

360K tokens/s

Background fill capacity

10B tokens/day

Background-inclusive replicas

4

Foreground avg util

60.1%

Foreground burst util

90.2%

Foreground throughput util

25.7%

Total utilization

31.3%

Agent trace load & capacity

Shows normal foreground load, burst demand, background load, and total concurrency capacity.

Average active tracesBurst active tracesBackground tracesTotal active workloadConcurrency capacity

Agent throughput (TPS) load & capacity

Shows live foreground throughput, scheduled background throughput, and total capacity.

Foreground TPSScheduled background TPSTotal workload TPSThroughput capacity

Next step

Want us to validate the model against your workload?

Bring your traces, token mix, and GPU assumptions. We can help translate this model into a deployment plan.

Talk to us

© 2026 Subconscious Systems Technologies, Inc.

Subconscious