Updated pricing · usage-20261019-v3

Jev and Closing Threads

Both services turn supplied text into structured decisions. Their prices, output types, and request limits differ.

Jev’s published input rate is about 10.4 times lower. Jev lists input at $0.042 per million tokens and no output-token charge. Closing Threads charges $0.4375 per million input tokens and $0.9375 per million output tokens. Private and Standard use the same rates, with no per-request minimum. These rates are lower than the previous Closing Threads prices and leave room above current provider costs; they do not match Jev’s price.

Check the current API guide for live rates.

Published input price per million tokens

The bars use one linear scale from $0 to $0.60 per million input tokens. Values appear next to each bar.

Input price per million tokens Linear bar chart from zero to 0.60 US dollars per million input tokens. Jev is 0.042 dollars. Closing Threads Standard and Private are each 0.4375 dollars. $0$0.15$0.30$0.45$0.60 JevClosing Threads StandardClosing Threads Private $0.042$0.4375$0.4375
USD per million input tokens. Private has the same rate as Standard. Closing Threads also charges for output tokens.

Product comparison

Published API features and limits
MeasureClosing ThreadsJev
Token priceStandard and Private: $0.4375 input and $0.9375 output per million tokens.$0.042 input per million tokens; output is free.
Request chargeNo per-request minimum. Charges use reported token usage and round up to the next US$0.000001.No output charge is listed; this comparison does not infer other request charges.
Typed outputsBoolean, choice, and bounded score.Choice, Score, and Noul.
ProbabilityReturns validated typed answers. It does not claim calibrated probabilities.Can return probabilities for decisions. A probability does not guarantee that an individual answer is correct.
Request sizeUp to 8 questions. The conservative total budget is 8,192 tokens for input and output combined.Up to 64,000 tokens per request and 32,000 per question. No separate question-count limit is listed within the request context.

Illustrative cost for 2,000 input and 8 output tokens

At the listed rates, rounded up to the next US$0.000001:

These estimates use the same token counts; they are not measured requests. Tokenizers and request overhead differ, so real requests may use different token counts. At these rates, this Closing Threads example costs about 10.5 times more than Jev’s. This comparison does not claim price parity.

A narrow latency check

A provider-only test of 100 synthetic follow-up and routing examples returned 100/100 expected answers. Median latency was 167 ms and p95 was 317 ms at concurrency two. This measures inference only, not a complete Closing Threads API request. We have not measured Jev latency, so this is not a speed comparison.