Daily series from each machine's ledger: Claude Code tokens by model, how much of the plan's cap was used, local tokens by purpose, the work moved off Claude, energy, and the tool output that entered Claude's context. A trend is a benchmark too: a change should show as a step.
Mac Studio M5 Ultra 96 GB
Daily, 2026-08-14 to 2026-11-11. Hover a point for its value; each chart has a table.
Claude Code tokens by model
claude-fable-5-1
claude-fable-5-2
claude-haiku-5-5
claude-opus-4-8
claude-opus-5
claude-opus-5-5
claude-sonnet-5-5
other
Table
day
claude-fable-5-1
claude-fable-5-2
claude-haiku-5-5
claude-opus-4-8
claude-opus-5
claude-opus-5-5
claude-sonnet-5-5
other
2026-09-13
--
--
--
10.6M
166.6M
--
--
408k
2026-09-14
--
--
--
--
181.4M
--
--
--
2026-09-16
--
--
--
--
11.8M
--
--
--
2026-09-17
--
--
--
--
76.7M
--
--
--
2026-09-18
--
--
--
--
249.1M
--
--
--
2026-09-19
--
--
--
--
253.6M
--
--
--
2026-09-20
--
--
--
--
174.5M
--
--
--
2026-09-21
--
--
--
--
292.2M
--
--
--
2026-09-22
--
--
--
--
113.5M
--
--
--
2026-09-25
--
--
--
--
31.3M
--
--
--
2026-09-27
--
--
--
--
--
--
--
28k
2026-09-29
--
--
--
--
1.9M
--
--
--
2026-09-30
--
--
--
--
269.6M
--
--
2.4M
2026-10-01
--
--
--
--
233.9M
--
--
--
2026-10-02
--
--
--
--
35.0M
246.7M
13.1M
790k
2026-10-03
--
--
--
--
--
76.7M
121.2M
913k
2026-10-04
--
--
--
--
--
193.0M
52.8M
--
2026-10-05
--
--
--
--
3.0M
311.7M
--
--
2026-10-06
--
--
--
--
--
1654.3M
--
--
2026-10-07
--
--
--
--
--
1003.5M
--
--
2026-10-08
--
--
--
--
--
238.2M
23.5M
--
2026-10-09
--
--
--
--
--
27.4M
12.3M
--
2026-10-10
43.4M
--
777k
--
--
102.1M
27.3M
--
2026-10-11
50.2M
--
2.9M
--
--
754.5M
2.0M
--
2026-10-12
6.6M
--
--
--
--
118.8M
--
--
2026-10-13
18.2M
--
--
197.6M
--
227.8M
--
--
2026-10-14
3.9M
--
195k
28.7M
--
17.4M
--
--
2026-10-15
42.0M
--
4.0M
36.5M
--
75.3M
456k
--
2026-10-16
6.9M
--
77k
69.5M
--
128.4M
--
--
2026-10-17
27.3M
--
--
--
--
587.3M
237k
--
2026-10-18
13.6M
--
--
--
--
168.1M
--
--
2026-10-19
--
--
--
--
--
416.3M
--
--
2026-10-20
15.6M
--
238k
16.6M
--
168.3M
--
--
2026-10-21
42.4M
--
--
--
--
223.7M
--
--
2026-10-22
--
--
--
--
--
589.7M
--
--
2026-10-23
--
--
316k
--
--
599.0M
--
--
2026-10-24
--
--
1.0M
--
--
383.6M
--
--
2026-10-25
--
--
2.9M
--
--
1977.3M
401k
--
2026-10-26
--
--
2.1M
--
--
1260.8M
2.2M
--
2026-10-27
--
--
1k
--
--
175.9M
--
--
2026-10-28
--
--
222k
--
--
517.4M
--
--
2026-10-29
--
--
18k
--
--
261.1M
--
--
2026-10-30
--
--
58k
--
--
666.1M
--
--
2026-10-31
--
--
92k
--
--
1177.1M
--
--
2026-11-01
--
--
--
--
57.5M
312.0M
--
--
2026-11-02
--
--
310k
--
--
155.5M
--
--
2026-11-03
--
--
711k
--
--
307.1M
--
--
2026-11-04
--
--
289k
--
1.9M
92.0M
--
--
2026-11-05
--
--
--
--
--
3.0M
--
--
2026-11-06
--
--
191k
--
--
95.3M
--
--
2026-11-07
--
--
2.5M
--
--
1308.0M
--
--
2026-11-08
--
--
1.6M
--
--
776.7M
--
--
2026-11-09
--
--
809k
--
--
343.2M
--
--
2026-11-10
--
--
3.8M
--
--
539.0M
--
--
2026-11-11
4.2M
25.9M
2.9M
--
--
1276.4M
--
--
Input, output and cache tokens from ccusage; the plan-cap chart is what the subscription counts.
Plan cap used
5-hour window, daily peak
7-day window, daily peak
Table
day
5-hour window, daily peak
7-day window, daily peak
2026-10-13
17%
59%
2026-10-14
5%
59%
2026-10-15
11%
7%
2026-10-16
12%
12%
2026-10-17
48%
28%
2026-10-18
6%
32%
2026-10-19
18%
39%
2026-10-20
10%
47%
2026-10-21
13%
47%
2026-10-22
24%
21%
2026-10-23
24%
36%
2026-10-24
10%
42%
2026-10-25
40%
64%
2026-10-26
23%
76%
2026-10-27
4%
79%
2026-10-28
15%
5%
2026-10-29
8%
9%
2026-10-30
24%
17%
2026-10-31
0%
31%
2026-11-01
--
36%
2026-11-02
--
39%
2026-11-03
--
42%
2026-11-04
--
2%
2026-11-05
--
2%
2026-11-06
--
4%
2026-11-07
--
21%
2026-11-08
--
30%
2026-11-09
--
35%
2026-11-10
--
44%
2026-11-11
--
19%
Peak of each day's statusline readings.
Local tokens by purpose
eval
internal
smoke
unknown
watchdog
work
work (alias mapping)
Table
day
eval
internal
smoke
unknown
watchdog
work
work (alias mapping)
2026-10-06
--
--
--
1.2M
--
--
--
2026-10-07
--
--
--
212k
--
--
--
2026-10-08
--
--
--
250k
--
--
36.7M
2026-10-09
--
--
--
94k
--
--
37.8M
2026-10-10
--
--
--
801k
--
--
2.8M
2026-10-11
--
--
--
1.4M
--
--
1.5M
2026-10-12
--
--
--
142k
--
--
--
2026-10-13
16k
--
2k
338k
--
3k
--
2026-10-14
74k
--
768
--
--
--
--
2026-10-15
261k
--
768
216
--
481k
--
2026-10-16
699k
--
768
--
--
1k
--
2026-10-17
--
--
--
--
--
0
--
2026-10-18
3.1M
--
2k
--
20k
192k
--
2026-10-19
1.4M
--
6k
--
15k
2k
--
2026-10-20
4k
--
2k
--
10k
5k
--
2026-10-21
337k
--
732
--
25k
7k
--
2026-10-22
1.5M
--
1k
0
9k
6k
--
2026-10-23
676k
--
2k
--
12k
8k
--
2026-10-24
37
1.5M
2k
51k
2k
--
--
2026-10-25
163k
--
4k
17k
3k
--
--
2026-10-26
9k
42k
756
185k
2k
--
--
2026-10-27
--
31k
--
1.3M
468
--
--
2026-10-28
86k
11.7M
--
--
2k
2k
--
2026-10-29
--
14.2M
--
--
627
--
--
2026-10-30
2.2M
641k
--
585k
1k
--
--
2026-10-31
3k
39k
1k
0
2k
--
--
2026-11-01
24
--
732
--
564
--
--
2026-11-02
1.8M
164k
--
--
1k
--
--
2026-11-03
291
76k
9k
--
2k
--
--
2026-11-04
2.6M
89k
5k
--
2k
--
--
2026-11-05
305k
76k
--
--
998
--
--
2026-11-06
61k
7k
--
--
519
--
--
2026-11-07
23k
23k
9k
--
512
--
--
2026-11-08
1.6M
277k
8k
--
--
--
--
2026-11-09
3.4M
130k
3k
--
210
--
--
2026-11-10
5.5M
25k
22k
--
54k
59k
--
2026-11-11
2.1M
27k
16k
0
65k
336k
--
Gateway tokens in and out. Work is what local displaced; eval and smoke are its cost.
Displaced tokens by lane
agent
code
decide
extract
unlaned
Table
day
agent
code
decide
extract
unlaned
2026-10-13
--
3k
--
--
--
2026-10-15
94k
--
259k
7k
121k
2026-10-16
--
67
--
--
1k
2026-10-17
--
--
--
--
0
2026-10-18
11k
--
179k
--
3k
2026-10-19
--
--
--
--
2k
2026-10-20
--
--
--
--
5k
2026-10-21
--
--
--
--
7k
2026-10-22
--
--
--
--
6k
2026-10-23
--
--
--
--
8k
2026-10-28
--
2k
--
--
472
2026-11-10
--
247
--
22k
37k
2026-11-11
--
7k
--
142k
187k
Tagged work served locally instead of by Claude, tokens in and out.
Energy, kWh per day
Table
day
kWh
2026-10-13
0.53 kWh
2026-10-14
0.10 kWh
2026-10-15
0.07 kWh
2026-10-16
0.23 kWh
2026-10-19
0.48 kWh
2026-10-20
0.33 kWh
2026-10-21
0.44 kWh
2026-10-22
1.04 kWh
2026-10-23
0.94 kWh
2026-10-24
1.31 kWh
2026-10-25
0.91 kWh
2026-10-26
0.55 kWh
2026-10-27
0.05 kWh
2026-10-28
0.50 kWh
2026-10-29
2.17 kWh
2026-10-30
0.77 kWh
2026-10-31
0.88 kWh
2026-11-01
0.30 kWh
2026-11-02
0.30 kWh
2026-11-03
0.47 kWh
2026-11-04
0.56 kWh
2026-11-05
0.24 kWh
2026-11-06
0.28 kWh
2026-11-07
0.68 kWh
2026-11-08
0.50 kWh
2026-11-09
0.45 kWh
2026-11-10
0.51 kWh
2026-11-11
0.54 kWh
macmon power, integrated per day; the dollar figure at the ledger's tariff is in each point's label. 2 days are left out because their CPU was not read (2026-10-17 .. 2026-10-18).
Tool-result tokens into Claude's context
Bash
Read
WebFetch
WebSearch
mcm-engine get_entry
mcm-engine report_error
mcm-engine search
other
Table
day
Bash
Read
WebFetch
WebSearch
mcm-engine get_entry
mcm-engine report_error
mcm-engine search
other
2026-08-16
13k
94k
53k
44k
--
--
25
8k
2026-08-17
14k
51k
4k
--
--
--
3k
8k
2026-08-18
247
--
--
--
--
--
--
--
2026-08-21
13k
2k
--
--
--
--
--
11k
2026-09-03
6k
1k
--
--
--
--
6k
2k
2026-09-06
65k
0
--
4k
--
--
2k
2k
2026-09-07
95k
0
831
5k
--
158
99k
13k
2026-09-08
104k
84
--
--
--
665
151k
5k
2026-09-09
60k
21
--
--
--
78
26k
1k
2026-09-10
3k
--
391
--
--
--
1k
0
2026-09-11
54k
--
--
--
--
117
35k
4k
2026-09-12
86k
--
--
--
--
282
123k
8k
2026-09-13
51k
37k
6k
533
713
111
73k
2k
2026-09-14
87k
--
--
--
--
331
100k
5k
2026-09-16
18k
--
--
--
--
39
16k
2k
2026-09-17
33k
--
--
--
--
156
56k
1k
2026-09-18
93k
33
--
--
--
551
111k
11k
2026-09-19
64k
4k
--
612
2k
549
67k
4k
2026-09-20
79k
--
978
--
--
571
81k
9k
2026-09-21
79k
0
--
--
2k
514
84k
5k
2026-09-22
13k
--
1k
5k
--
78
9k
1k
2026-09-25
24k
--
--
--
2k
39
40k
3k
2026-09-29
6k
--
--
--
--
--
6k
335
2026-09-30
112k
6k
--
--
--
267
133k
12k
2026-10-01
53k
686
--
--
--
118
44k
7k
2026-10-02
258k
4k
21k
--
--
112
221k
9k
2026-10-03
521k
11k
--
--
--
78
203k
17k
2026-10-04
263k
5k
--
--
--
158
97k
19k
2026-10-05
234k
2k
270
--
--
509
180k
7k
2026-10-06
2.5M
185k
748
--
2k
3k
1.5M
96k
2026-10-07
1.4M
49k
--
829
3k
2k
1.2M
69k
2026-10-08
548k
13k
--
--
771
1k
409k
39k
2026-10-09
94k
2k
--
--
--
39
70k
22k
2026-10-10
278k
10k
--
8k
--
192
221k
44k
2026-10-11
2.9M
119k
1k
3k
11k
5k
1.4M
80k
2026-10-12
328k
23k
--
--
--
436
263k
8k
2026-10-13
368k
100k
8k
991
1k
1k
408k
25k
2026-10-14
77k
18k
4k
--
3k
177
151k
5k
2026-10-15
186k
57k
122k
6k
7k
44k
228k
21k
2026-10-16
324k
163k
3k
2k
4k
84k
346k
17k
2026-10-17
1.2M
451k
3k
845
15k
88k
999k
38k
2026-10-18
287k
102k
4k
719
8k
31k
290k
14k
2026-10-19
768k
168k
--
--
40k
85k
834k
23k
2026-10-20
284k
98k
2k
2k
10k
28k
366k
16k
2026-10-21
266k
121k
53
1k
4k
28k
235k
29k
2026-10-22
715k
2.9M
--
--
11k
54k
561k
32k
2026-10-23
1.3M
1.8M
6k
11k
15k
53k
564k
40k
2026-10-24
379k
133k
15k
33k
12k
26k
386k
14k
2026-10-25
1.4M
580k
21k
78k
41k
141k
1.5M
115k
2026-10-26
759k
324k
57k
5k
39k
55k
855k
61k
2026-10-27
45k
3k
--
--
--
4k
22k
2k
2026-10-28
311k
87k
18k
6k
8k
25k
391k
27k
2026-10-29
240k
72k
--
811
7k
24k
262k
15k
2026-10-30
749k
535k
731
3k
76k
52k
833k
49k
2026-10-31
1.1M
350k
--
4k
58k
72k
1.0M
65k
2026-11-01
263k
84k
--
--
12k
18k
243k
23k
2026-11-02
206k
100k
39k
6k
8k
10k
153k
10k
2026-11-03
257k
103k
--
1k
8k
27k
256k
20k
2026-11-04
114k
51k
--
--
4k
7k
157k
10k
2026-11-05
2k
--
--
--
613
--
7k
1k
2026-11-06
164k
32k
--
--
4k
20k
161k
8k
2026-11-07
745k
228k
8k
--
36k
51k
913k
65k
2026-11-08
631k
215k
--
--
22k
54k
754k
52k
2026-11-09
481k
145k
--
--
8k
41k
523k
27k
2026-11-10
593k
293k
36k
52k
21k
71k
732k
55k
2026-11-11
1.5M
670k
14k
691
36k
116k
1.9M
104k
Estimated as characters / 4 per tool result (soh context). Marked: 2026-10-13: rtk hook and bashOutputMaxChars 8000; 2026-10-14: mcm-engine 3.12 search summaries.
Delegation acceptance by lane
code
extract
Table
day
code
extract
2026-10-15
--
1.00
2026-10-28
1.00
1.00
2026-11-10
--
0.80
2026-11-11
0.60
1.00
Share of closed outcomes, reported or inferred from the caller's transcript, that were accepted rather than rejected.