Admin

What NephAI has been doing

The work NephAI carried out for you over the last 90 days, and what it cost.

15 kinds of work ran

What it cost
$337.25
Times it thought something through
108,764
Amount of text handled
454,747,533
Average cost each time
$0.0031

Where the work went

15 in use
Where the work went with calls, tokens, cost, and share of spending
Kind of workCallsTokensCostShare
Automation reasoning

Plans and verifies the next covered action from current evidence.

9,03575,903,066$114.8434.1%
Thread sorting

Separates Church, work, and personal conversations.

23,96116,708,830$62.8618.6%
Inbox understanding

Reads each conversation and works out what it needs.

13,746240,410,814$53.8516.0%
Automation follow-through

Confirms from current evidence when prepared Automation work has finished.

1,51327,348,626$40.6312.0%
Contact profiles

Learns enough about a new correspondent to prepare work correctly.

57,64470,725,216$21.466.4%
Prepared replies

Writes the canonical reply waiting in a conversation.

1,08512,280,839$16.644.9%
Ask NephAI

Answers your questions, gathers context, and prepares guarded actions.

2325,662,876$11.463.4%
Handbook fact-checking

Checks a prepared Church draft against retrieved Handbook passages.

4602,077,599$6.031.8%
Handbook lookup

Plans the Handbook questions and passages needed for a response.

7141,846,743$5.031.5%
Draft checking

Checks every prepared reply against your own words and calendar before you see it.

3111,348,460$3.731.1%
Stewardship review

Reviews your stewardships and prepares the next useful move.

48369,149$0.57980.2%
Recipient insights

Reads your history with the compose recipients into a short pre-writing brief.

931,679$0.08990.0%
Composer

Writes, refines, and reshapes email while you compose.

222,394$0.03310.0%
Draft instructions

Reads what you type in the composer so Draft does the right thing with it.

37,931$0.01950.0%
Public-web research

Finds current public sources and keeps cited evidence for Ask NephAI.

13,311$0.00840.0%

Which model did it

9 in use
Which model did it with calls, tokens, cost, and share of spending
ModelCallsTokensCostShare
GPT 5.6 Terra11,698118,455,306$173.2660.5%
Kimi K323,96516,743,104$63.0322.0%
Gemini 3.1 Flash Lite57,64470,725,216$21.467.5%
Grok 4.51,4945,304,481$14.885.2%
Claude Opus 51492,355,231$8.112.8%
Claude Opus 4.830488,024$3.191.1%
Claude Opus 4.721361,622$1.960.7%
GLM 5.244334,875$0.40650.1%
Gemini 2.5 Flash13,311$0.00840.0%