Jev vs GPT-5.6 Luna: cost per decision
Luna is the cheapest frontier-family LLM most teams reach for when classifying. Here is what the same workload costs on each, with output tokens counted.
| Scenario | Jev 1.13 | GPT-5.6 Luna | Ratio |
|---|---|---|---|
Light 10k items/day, 300-token state, 2 questions | $5.29 | $36.00 | 6.8x |
Production 100k items/day, 500-token state, 4 questions | $93.24 | $624.00 | 6.7x |
High volume 2M items/day, 300-token state, 1 question | $856.80 | $4,800 | 5.6x |
Monthly, 30 days. Both sides answer all questions for an item in a single call. GPT-5.6 Luna additionally pays for its output tokens at $1.2/M; Jev 1.13 has no output charge. List prices at standard tier, excluding caching and batch discounts.
Where GPT-5.6 Luna still wins
Luna can write the answer. If the step after classification needs prose, a summary or a tool call with free-form arguments, Jev cannot produce it.
The honest caveat
This compares price, not accuracy. A model that is 100x cheaper and wrong 5% more often can easily be the more expensive choice once you price the mistakes. Jev returns a calibrated probability with every answer, so the practical move is to route the confident cases automatically and send the rest to GPT-5.6 Luna or a human — which makes this a routing decision rather than a replacement decision.