List individual B.AI usage records.
wallet-cli bai usage-records [--limit <n>] [--offset <n>] [--sort <asc|desc>] [--cursor <c>] [options]
Lists the API key's usage, one record per request: model, tokens, credits, and latency. No wallet or password is needed.
Page with --limit / --offset, or pass the nextCursor of one page back with --cursor to get the following page.
Requires a B.AI API key stored with config baiApiKey; see bai.
| Option | Description |
|---|---|
--limit <n> |
Rows per page, 1–200 (default 20) |
--offset <n> |
Rows to skip (default 0); must be a multiple of --limit, because B.AI pages by page number |
--sort <asc|desc> |
Order by creation time (default desc) |
--cursor <c> |
nextCursor from the previous page |
Plus the global options.
wallet-cli bai usage-records --limit 3B.AI usage records
records:
1.
id: usg_8f3a2c71
created At: 2026-09-15T08:21:44.000Z
model: gpt-5-mini
input Tokens: 1840
output Tokens: 612
total Tokens: 2452
credits: 1200
latency Ms: 2,840
source: api
2.
id: usg_8f3a2b09
created At: 2026-09-15T08:19:02.000Z
model: claude-sonnet-5
input Tokens: 3210
output Tokens: 958
total Tokens: 4168
credits: 4600
latency Ms: 5,120
source: api
3.
id: usg_8f39f6d4
created At: 2026-09-14T13:05:37.000Z
model: gpt-5-mini
input Tokens: 920
output Tokens: 301
total Tokens: 1221
credits: 600
latency Ms: 1,730
source: api
pagination:
offset: 0
limit: 3
total: Not available
has More: Yes
next Cursor: eyJpZCI6InVzZ184ZjM5ZjZkNCJ9wallet-cli bai usage-records --limit 3 -o json{"schema":"wallet-cli.result.v1","success":true,"command":"bai.usage-records","data":{"records":[{"id":"usg_8f3a2c71","createdAt":"2026-09-15T08:21:44.000Z","model":"gpt-5-mini","inputTokens":"1840","outputTokens":"612","totalTokens":"2452","credits":"1200","latencyMs":2840,"source":"api"},{"id":"usg_8f3a2b09","createdAt":"2026-09-15T08:19:02.000Z","model":"claude-sonnet-5","inputTokens":"3210","outputTokens":"958","totalTokens":"4168","credits":"4600","latencyMs":5120,"source":"api"},{"id":"usg_8f39f6d4","createdAt":"2026-09-14T13:05:37.000Z","model":"gpt-5-mini","inputTokens":"920","outputTokens":"301","totalTokens":"1221","credits":"600","latencyMs":1730,"source":"api"}]},"meta":{"durationMs":1323,"warnings":[],"pagination":{"offset":0,"limit":3,"total":null,"hasMore":true,"nextCursor":"eyJpZCI6InVzZ184ZjM5ZjZkNCJ9"}}}Records are newest first. credits is what each request cost, and latencyMs how long it took. hasMore: true (text has More: Yes) means there are more records: fetch the next page with --offset 3, or with --cursor eyJpZCI6InVzZ184ZjM5ZjZkNCJ9.
| Field | Type | Meaning |
|---|---|---|
records[] |
array | Usage records, below |
Paging is in meta.pagination: offset, limit, total (always null, since B.AI returns no count), and when B.AI supplies them hasMore and nextCursor. hasMore: false means this is the last page.
Each record (a field B.AI does not supply is absent):
| Field | Type | Meaning |
|---|---|---|
id |
string | Record id |
createdAt |
string | When the request was made |
model |
string | Model used |
inputTokens / outputTokens / totalTokens |
string | Token counts |
credits |
string | Credits charged |
latencyMs |
number | Request duration, in milliseconds |
source |
string | Where the request came from |
0 success · 1 execution failure (bai_auth_failed; bai_rejected; provider_error — including a --cursor B.AI does not accept; timeout) · 2 usage error (bai_credentials_missing; invalid_value — --offset not a multiple of --limit, or out of range).