@alexeiled/claude-router 1.17.0 → 1.18.0

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
@@ -1,7 +1,7 @@
1
1
  {
2
2
  "name": "router",
3
3
  "displayName": "Router",
4
- "version": "1.17.0",
4
+ "version": "1.18.0",
5
5
  "description": "Native model and effort routing within your provider. Anthropic, DeepSeek and z.ai defaults, custom provider profiles, classifier advice and one model per subagent.",
6
6
  "author": {
7
7
  "name": "Alexei Ledenev",
package/CHANGELOG.md CHANGED
@@ -1,5 +1,27 @@
1
1
  # Changelog
2
2
 
3
+ ## 1.18.0 (2026-10-30)
4
+
5
+ The band and the pane now show one session cost readout: the same number, the same words, at the same moment.
6
+
7
+ ### Changed
8
+
9
+ - The band's second row shows the answer and the parts that add up to it: `session routing +$0.18 (+4%) vs Opus 5.5 · high = −$0.42 cheaper models +$0.60 cache rebuilds · 34 replies`. Before, it showed `extra $0.18` beside only the largest cost (`cold-cache writes +$0.60`), which did not add up. On a narrow band the parts lose their words (`= −$0.42 + $0.60`), then the reply count drops, then the parts.
10
+ - The Now tab's SESSION block starts with the band's row and the ledger that adds up to it, with **Usage** beside it. When cache rebuilds cost more than the cheaper models saved, `▲ switches cost more than they saved` follows with **Policy**. The cost line at the end of the block is gone.
11
+ - One vocabulary everywhere: `routing −$x (−y%) vs <model>`, `cheaper models`, `stronger models`, `cache rebuilds`. The Usage tab, `/router`, TODAY and SUBAGENTS use it too. The policy's estimate for the switch it weighs is now `next switch ≈ $x`, so it no longer reads like the cache rebuilds already paid.
12
+ - After a `/model` change the band names the models the session's replies ran against, as the Usage tab does, instead of repricing the session against the new model.
13
+
14
+ ### Fixed
15
+
16
+ - The pane counts the replies of the turn still running, marked `(N live)`. Before, it counted a turn only when the turn ended, so during a long turn the band showed a cost the pane did not have.
17
+ - The band's session readout no longer starts over when Claude Code restarts a resumed session: it reads the session's usage record, as the pane does.
18
+
19
+ ## 1.17.1 (2026-10-28)
20
+
21
+ ### Fixed
22
+
23
+ - When the conversation fills more than 80% of every route's context window, the band and panel say `! context full` and the router keeps Claude's model. Before, they said `! model unavailable`, which read as an outage or an `availableModels` block. `model unavailable` now means only that: a model the allowlist blocks, one that overflowed its window, or one that cannot take images.
24
+
3
25
  ## 1.17.0 (2026-10-28)
4
26
 
5
27
  Route models and effort within your configured provider. Anthropic remains the default. DeepSeek and z.ai gain built-in profiles, and other Anthropic-compatible providers can use custom profiles without another proxy.
@@ -133,7 +133,7 @@ Nothing is recorded with `activityRouting: off`. Stats run inside a guard: a fai
133
133
 
134
134
  ### Routing vs your model
135
135
 
136
- After each main reply, where the response is observed, `savings.mjs` prices the reply on the model that served it and the same tokens on "your model": the session's model and the effort Claude Code sent on the turn's first request. `compareModels` prices each reply against every model in `router.json` as "your model", at the session effort. Each keeps a simulated single cache in the loop under its models key in `yoursBy` (`{ total, at, onYours, model, effort }`), which a history reset clears; after an effort change the next reply finds it cold. The session's totals per model are `savingsBy` in the view, and `updateView` sets `savings` to the entry for the model `/model` selects whenever the model or the totals change, so a `/model` change re-prices the whole session without a replay. Each reply also becomes a `replyEntry` in `usage.mjs`: its time, its cell `activity|tier|modelId@effort` with the model that served it, its prices against the model selected then, its input, cache-read and output tokens, and what the same tokens cost on each model as yours. Models are kept by id, not by models key, because a key can later name another model. At turn completion `recordUsage` adds the turn's entries to the session's own `$.store` key `usage:v1:<session id>`, by local calendar day: per day the cells' nine sums, the what-if sums, the replies per your model, the routed cost per hour, the turns, and the route switches by tier and by activity, with `since`, the record's first write. Days older than 400 days leave the record when it is written. Each key has one writer, and Claude Code 2.1.296 writes one key under a file lock (observed in its code; the `$.store` types do not promise it), so sessions that finish turns at the same moment do not overwrite each other. The view's `usageStore` adds up two parts: every other session's key, read at session start and when the pane opens, and this session's record (`usageSession`), which the router reads once, holds in memory, adds each finished turn to and then writes, so a pane opened during that write counts the turn once. On that read, `prunedUsage` names other sessions' records with no day in the last 400 days and, past 1,000 records, those whose last day is oldest; a record that does not read whole, such as one a later version wrote, is never named. Each named record is read again and deleted only if it has not changed, so a session that wrote its record since keeps it; with no conditional delete in `$.store`, a write between that read and the delete is still lost. A day that does not read is dropped, and the others stay. The Usage tab sums the days of the chosen period (`summarize`); the period is kept across sessions in `usage:period`. Before these records each session kept undated totals in `savings:v1:<session id>`; they are no longer written, and the view's `savingsStore`, their sum, is added to the All period. **Delete history** first writes the time in `savings:reset:v1`, then deletes every usage, savings, activity stats, metrics and subagent record. A record whose first reply is older than that time reads as empty, and each session reads the time before it adds a turn, so a session that still holds its record from before a Reset in another session drops it instead of writing it back. The comparison is by milliseconds: a record whose first reply has the reset's millisecond, or one written while the clock had moved back, still counts. A stored value of any other shape reads as empty. A reply with routing off or after an engine fallback is not counted, but it still moves your model's cache. A reply whose cache lifetime cannot be read loses its prices only, and your model's cache still follows it; a usage without rate limits prices writes at five minutes. No failure here fails the turn. [Evaluation](evaluation.md#routing-vs-your-model) gives the method and its limits.
136
+ After each main reply, where the response is observed, `savings.mjs` prices the reply on the model that served it and the same tokens on "your model": the session's model and the effort Claude Code sent on the turn's first request. `compareModels` prices each reply against every model in `router.json` as "your model", at the session effort. Each keeps a simulated single cache in the loop under its models key in `yoursBy` (`{ total, at, onYours, model, effort }`), which a history reset clears; after an effort change the next reply finds it cold. Each reply also becomes a `replyEntry` in `usage.mjs`: its time, its cell `mode|activity|tier|modelId@effort` with the model that served it, its prices against the model selected then, its input, cache-read and output tokens, and what the same tokens cost on each model as yours. Models are kept by id, not by models key, because a key can later name another model. Until the turn ends its entries are also the view's `usageLive`; every readout (the two-row band, the Now and Usage tabs, `/router`) adds them to the record with `withReplies`, so they count the same replies at the same moment. At turn completion the entries leave `turnUsage` in the same tick they enter the held record, and every view write of the records carries `usageLive` with them, so a pane opened mid-write, a failed store write or a Reset counts a turn once or not at all, never twice. At turn completion `recordUsage` adds the turn's entries to the session's own `$.store` key `usage:v1:<session id>`, by local calendar day: per day the cells' nine sums, the what-if sums, the replies per your model, the routed cost per hour, the turns, and the route switches by tier and by activity, with `since`, the record's first write. Days older than 400 days leave the record when it is written. Each key has one writer, and Claude Code 2.1.296 writes one key under a file lock (observed in its code; the `$.store` types do not promise it), so sessions that finish turns at the same moment do not overwrite each other. The view's `usageStore` adds up two parts: every other session's key, read at session start and when the pane opens, and this session's record (`usageSession`), which the router reads once, holds in memory, adds each finished turn to and then writes, so a pane opened during that write counts the turn once. On that read, `prunedUsage` names other sessions' records with no day in the last 400 days and, past 1,000 records, those whose last day is oldest; a record that does not read whole, such as one a later version wrote, is never named. Each named record is read again and deleted only if it has not changed, so a session that wrote its record since keeps it; with no conditional delete in `$.store`, a write between that read and the delete is still lost. A day that does not read is dropped, and the others stay. The Usage tab sums the days of the chosen period (`summarize`); the period is kept across sessions in `usage:period`. Before these records each session kept undated totals in `savings:v1:<session id>`; they are no longer written, and the view's `savingsStore`, their sum, is added to the All period. **Delete history** first writes the time in `savings:reset:v1`, then deletes every usage, savings, activity stats, metrics and subagent record. A record whose first reply is older than that time reads as empty, and each session reads the time before it adds a turn, so a session that still holds its record from before a Reset in another session drops it instead of writing it back. The comparison is by milliseconds: a record whose first reply has the reset's millisecond, or one written while the clock had moved back, still counts. A stored value of any other shape reads as empty. A reply with routing off or after an engine fallback is not counted, but it still moves your model's cache. A reply whose cache lifetime cannot be read loses its prices only, and your model's cache still follows it; a usage without rate limits prices writes at five minutes. No failure here fails the turn. [Evaluation](evaluation.md#routing-vs-your-model) gives the method and its limits.
137
137
 
138
138
  ### Subagents
139
139
 
@@ -153,13 +153,13 @@ The Usage tab compares what routed replies cost with what the same tokens would
153
153
  Method, per main reply that routing chose:
154
154
 
155
155
  - **Routed** prices the usage Claude reported (uncached input, cache read, cache write, output) on the model that served it.
156
- - **Yours** prices the same tokens on each model in `router.json` at the session effort. The Usage tab keeps each reply against the model selected when it ran, and **What if** shows the period against each model alone; the two-row band and `/router` show the session against the model `/model` selects now, so a `/model` change shows the whole session against the new model there. While this reply and the one before it ran on that model's route, it is the reply's own usage, so a session on your model differs by exactly $0.00. Otherwise each model keeps one cache that never switched: it reads the previous prompt, up to this one less its uncached input, inside the cache lifetime, unless a compaction shrank the history or, on a model that keeps one cache per effort, the effort changed since, and never less than the API actually read; it writes the rest. Output tokens are the same.
156
+ - **Yours** prices the same tokens on each model in `router.json` at the session effort. The Usage tab keeps each reply against the model selected when it ran, and **What if** shows the period against each model alone; the two-row band, the Now tab and `/router` show the same session readout, the running turn's replies included. While this reply and the one before it ran on that model's route, it is the reply's own usage, so a session on your model differs by exactly $0.00. Otherwise each model keeps one cache that never switched: it reads the previous prompt, up to this one less its uncached input, inside the cache lifetime, unless a compaction shrank the history or, on a model that keeps one cache per effort, the effort changed since, and never less than the API actually read; it writes the rest. Output tokens are the same.
157
157
  - **Parts**: the served model on your cache against yours is the model price, under cheaper models when negative and stronger than yours when positive; routed against the served model on your cache is the cache the switches cost, never negative. They sum to the difference. The pane rounds to cents and gives the remainder only to a part that is not zero: a surplus to the switch part, else the stronger part, else the cheaper part toward $0.00; a shortfall comes off the stronger part, down to $0.00, then the cheaper part. The shown parts add up, none changes sign, and none appears from rounding alone.
158
158
  - **Prices**: `router.json` list prices through `ratesAt`, so Haiku 5.5 above 100,000 prompt tokens is priced at 5x. Claude Code reports no 5m/1h split to a Mod, so cache writes use the lifetime Claude Code 2.1.296 gives the main conversation by its own rule: one hour on a Claude plan unless an environment variable or the `promptCacheTtl` setting says otherwise, five minutes with an API key.
159
159
 
160
160
  Checks:
161
161
 
162
- - Unit tests (`test/savings.test.mjs`) cover a session that saves, one that costs more with Haiku as your model, equal routes at exactly 0, cache expiry at 5m and 1h, a compaction, a change of your model or effort, the 100K surcharge, an unpriced model, and a corrupted store. `test/hook-routing.test.mjs` covers two sessions finishing turns at the same moment, and a `/model` change mid-session that shows the whole session against the new model. `test/usage.test.mjs` covers calendar days across month ends and daylight saving changes, the periods, the record cap and expiry, the trend buckets and what is worth a look; `test/usage-view.test.mjs` covers the Usage tab.
162
+ - Unit tests (`test/savings.test.mjs`) cover a session that saves, one that costs more with Haiku as your model, equal routes at exactly 0, cache expiry at 5m and 1h, a compaction, a change of your model or effort, the 100K surcharge, an unpriced model, and a corrupted store. `test/hook-routing.test.mjs` covers two sessions finishing turns at the same moment, a `/model` change mid-session that keeps each reply against its own model, and a running turn's replies counted once before and after the turn is recorded. `test/usage.test.mjs` covers calendar days across month ends and daylight saving changes, the periods, the record cap and expiry, the trend buckets and what is worth a look; `test/usage-view.test.mjs` covers the Usage tab.
163
163
  - A one-off replay, not checked in, of this machine's main-conversation transcripts from 2026-10-01 to 2026-10-10 (nine sessions with 10 or more replies) gave $0.00 for every session that ran only on the model it was compared with: three on Opus 5.5, one on Haiku 5.5. Every main-conversation cache write in them was a one-hour write, which is what the lifetime rule gives on this Max plan.
164
164
  - A live Claude Code 2.1.296 session on 2026-10-10, Opus 5.5 at `high` with turns pinned to Haiku 5.5, showed **Routed replies $0.48** beside **Cost $0.480 reported by Claude** for the same 16 replies; its last Haiku reply had a 114K-token prompt, which agrees only at the 5x rate. A Haiku 5.5 session with a pinned Opus turn showed $0.45 routed against $0.446 reported.
165
165
 
@@ -39,7 +39,7 @@ The status band is one line above the prompt: a tier meter, the activity, the ro
39
39
  | Classifier support | Per-tier probabilities from the last classification; under **Why**, the support for the switch against the bar the policy required, and the switch cost. Pins and skipped calls leave it empty. |
40
40
  | Activity | The classifier's activity choice as a verb with its share, the two next-best answers, and the threshold at which an activity applies; `uncertain` shown dim. Empty with routing off, a pin, or no answer. |
41
41
  | Session and today (Now) | Every main-conversation reply since Claude Code started the session as a strip colored by tier, pairs of cells merging past 48 cells or on a narrow pane; a dot where the router did not serve, and `since HH:MM` once a restart started the strip again. The replies per tier and per activity, the route switches and the last one, and routing against your model for the session and, once another session added to it, for today; today's first finding worth a look. |
42
- | Routing vs your model (Usage) | For a period (this session, today, 7 or 30 days, this month, or all), routed replies at configured list prices against the same tokens on your model, the model `/model` selected at the session effort when each reply ran: a ledger from your model through cheaper models, stronger models and switch cache writes to routed; tokens and cache hit. `early sample` below 10 replies. Output length is held the same; answer quality is not measured. |
42
+ | Routing vs your model (Usage) | For a period (this session, today, 7 or 30 days, this month, or all), routed replies at configured list prices against the same tokens on your model, the model `/model` selected at the session effort when each reply ran: a ledger from your model through cheaper models, stronger models and cache rebuilds to routed. The two-row band and the Now tab's SESSION block show the session's answer and parts in the same words, the running turn's replies included; tokens and cache hit. `early sample` below 10 replies. Output length is held the same; answer quality is not measured. |
43
43
  | Breakdown (Usage) | The period's replies, cost and difference against your model by activity, model or tier, each row opening to its routes or activities; the cost per hour, day or week, or **What if**, the same tokens on each model alone; from 20 replies, the cells worth a look, linked to Routes or Policy. |
44
44
  | Across sessions (Classifier) | The counts kept since the last reset: labelled turns, tool agreement on the classifier's raw answer, overall and for `code`, `ops` and `explore`, the two largest mismatches, activity moves inside a tier taken and refused, and escalations after a cheaper activity move. The active classifier's p95 latency shows on its **Deadline** line. `/router` text shows the agreement. |
45
45
  | Cost | Claude Code's native usage value. A plan's list-price estimates are not subscription cash charges. |
@@ -47,12 +47,10 @@
47
47
  <text x="765.0" y="121" fill="#87d787" xml:space="preserve">100%</text>
48
48
  <text x="811.8" y="121" fill="#d0d0d0" xml:space="preserve">Router</text>
49
49
  <text x="24.0" y="140" fill="#7c7c7c" xml:space="preserve">session </text>
50
- <text x="94.2" y="140" fill="#6fcf97" font-weight="700" xml:space="preserve">saved $1.84 (22%)</text>
51
- <text x="226.8" y="140" fill="#d0d0d0" xml:space="preserve"> vs Opus 5.5 · xhigh</text>
52
- <text x="382.8" y="140" fill="#7c7c7c" xml:space="preserve"> · </text>
53
- <text x="421.8" y="140" fill="#7c7c7c" xml:space="preserve">64 replies</text>
54
- <text x="499.8" y="140" fill="#7c7c7c" xml:space="preserve"> · </text>
55
- <text x="538.8" y="140" fill="#7c7c7c" xml:space="preserve">cold-cache writes +$0.47</text>
50
+ <text x="94.2" y="140" fill="#6fcf97" font-weight="700" xml:space="preserve">routing −$1.84 (−22%)</text>
51
+ <text x="258.0" y="140" fill="#d0d0d0" xml:space="preserve"> vs Opus 5.5 · xhigh</text>
52
+ <text x="414.0" y="140" fill="#7c7c7c" xml:space="preserve"> = −$2.31 + $0.47</text>
53
+ <text x="562.2" y="140" fill="#7c7c7c" xml:space="preserve"> · 64 replies</text>
56
54
  <text x="24.0" y="159" fill="#7c7c7c" xml:space="preserve">pin next turn </text>
57
55
  <rect x="133.2" y="145" width="70.2" height="18" rx="4" fill="#2e3340"/>
58
56
  <text x="133.2" y="159" fill="#d0d0d0" xml:space="preserve">[ micro ]</text>
@@ -150,11 +148,9 @@
150
148
  <text x="297.0" y="484" fill="#7c7c7c" xml:space="preserve"> · </text>
151
149
  <text x="336.0" y="484" fill="#7c7c7c" xml:space="preserve">= fits</text>
152
150
  <text x="382.8" y="484" fill="#7c7c7c" xml:space="preserve"> · </text>
153
- <text x="421.8" y="484" fill="#7c7c7c" xml:space="preserve">cost (session)</text>
154
- <text x="531.0" y="484" fill="#7c7c7c" xml:space="preserve"> · </text>
155
- <text x="570.0" y="484" fill="#d0d0d0" xml:space="preserve">ctx </text>
156
- <text x="601.2" y="484" fill="#87d787" xml:space="preserve">40%</text>
157
- <text x="624.6" y="484" fill="#d0d0d0" xml:space="preserve"> · cache hit </text>
158
- <text x="726.0" y="484" fill="#87d787" xml:space="preserve">100%</text>
159
- <text x="772.8" y="484" fill="#d0d0d0" xml:space="preserve">Router</text>
151
+ <text x="421.8" y="484" fill="#d0d0d0" xml:space="preserve">ctx </text>
152
+ <text x="453.0" y="484" fill="#87d787" xml:space="preserve">40%</text>
153
+ <text x="476.4" y="484" fill="#d0d0d0" xml:space="preserve"> · cache hit </text>
154
+ <text x="577.8" y="484" fill="#87d787" xml:space="preserve">100%</text>
155
+ <text x="624.6" y="484" fill="#d0d0d0" xml:space="preserve">Router</text>
160
156
  </svg>
@@ -1,12 +1,12 @@
1
- <svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 665 629" role="img" aria-labelledby="title desc">
1
+ <svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 665 705" role="img" aria-labelledby="title desc">
2
2
  <title id="title">Router pane, Now tab</title>
3
3
  <desc id="desc">What runs next turn: tier, model and activity, with the reason and the classifier support it weighed; one pin button per tier with its support; this session since its first reply, as a strip of every reply colored by tier, with the replies per tier and per activity, the last route switch and what routing saved against your model; today across sessions; this session’s subagents as squares by model, one running, with what routing saved on them; today’s first finding worth a look with a link to Routes; and context, cache and cost.</desc>
4
4
  <style>
5
5
  text { font-family: ui-monospace, SFMono-Regular, Menlo, Consolas, 'Liberation Mono', monospace; font-size: 13px; }
6
6
  .cap { font-family: system-ui, -apple-system, 'Segoe UI', sans-serif; font-size: 12px; fill: #9a988f; }
7
7
  </style>
8
- <rect width="665" height="629" rx="8" fill="#1c1c1c"/>
9
- <rect x="10" y="10" width="645" height="609" rx="6" fill="none" stroke="#3f5f8f"/>
8
+ <rect width="665" height="705" rx="8" fill="#1c1c1c"/>
9
+ <rect x="10" y="10" width="645" height="685" rx="6" fill="none" stroke="#3f5f8f"/>
10
10
  <text x="24.0" y="34" fill="#5fafd7" font-weight="700" xml:space="preserve">ROUTING</text>
11
11
  <rect x="94.2" y="20" width="62.4" height="18" rx="4" fill="#87afff"/>
12
12
  <text x="94.2" y="34" fill="#0d1016" font-weight="700" xml:space="preserve">[ ◉ on ]</text>
@@ -35,7 +35,7 @@
35
35
  <text x="414.0" y="110" fill="#7c7c7c" xml:space="preserve"> applies at 60%+</text>
36
36
  <text x="24.0" y="129" fill="#d0d0d0" xml:space="preserve">Why </text>
37
37
  <text x="109.8" y="129" fill="#d0d0d0" xml:space="preserve">clear need for a stronger model</text>
38
- <text x="24.0" y="148" fill="#7c7c7c" xml:space="preserve"> Jev gave stronger tiers 88% (needs 82%) · switch ≈ $0.48</text>
38
+ <text x="24.0" y="148" fill="#7c7c7c" xml:space="preserve"> Jev gave stronger tiers 88% (needs 82%) · next switch ≈ $0.48</text>
39
39
  <text x="24.0" y="167" fill="#d0d0d0" xml:space="preserve">Mode </text>
40
40
  <text x="109.8" y="167" fill="#d0d0d0" xml:space="preserve">balanced (default)</text>
41
41
  <text x="24.0" y="186" fill="#d0d0d0" xml:space="preserve">Pin next </text>
@@ -50,120 +50,130 @@
50
50
  <text x="523.2" y="186" fill="#7c7c7c" xml:space="preserve"> next turn only</text>
51
51
  <text x="24.0" y="224" fill="#5fafd7" font-weight="700" xml:space="preserve">SESSION</text>
52
52
  <text x="78.6" y="224" fill="#7c7c7c" xml:space="preserve"> · since 14:18 · 86 routed replies · 9 route switches</text>
53
- <text x="39.6" y="243" fill="#5f9fe0" font-weight="700" xml:space="preserve">█</text>
54
- <text x="47.4" y="243" fill="#5f9fe0" font-weight="700" xml:space="preserve">█</text>
55
- <text x="55.2" y="243" fill="#5f9fe0" font-weight="700" xml:space="preserve">█</text>
56
- <text x="63.0" y="243" fill="#d7af5f" font-weight="700" xml:space="preserve">█</text>
57
- <text x="70.8" y="243" fill="#5f9fe0" font-weight="700" xml:space="preserve">█</text>
58
- <text x="78.6" y="243" fill="#5f9fe0" font-weight="700" xml:space="preserve">█</text>
59
- <text x="86.4" y="243" fill="#5f9fe0" font-weight="700" xml:space="preserve">█</text>
60
- <text x="94.2" y="243" fill="#5f9fe0" font-weight="700" xml:space="preserve">█</text>
61
- <text x="102.0" y="243" fill="#5f9fe0" font-weight="700" xml:space="preserve">█</text>
62
- <text x="109.8" y="243" fill="#e0708a" font-weight="700" xml:space="preserve">█</text>
63
- <text x="117.6" y="243" fill="#e0708a" font-weight="700" xml:space="preserve">█</text>
64
- <text x="125.4" y="243" fill="#e0708a" font-weight="700" xml:space="preserve">█</text>
65
- <text x="133.2" y="243" fill="#e0708a" font-weight="700" xml:space="preserve">█</text>
66
- <text x="141.0" y="243" fill="#5f9fe0" font-weight="700" xml:space="preserve">█</text>
67
- <text x="148.8" y="243" fill="#5f9fe0" font-weight="700" xml:space="preserve">█</text>
68
- <text x="156.6" y="243" fill="#5f9fe0" font-weight="700" xml:space="preserve">█</text>
69
- <text x="164.4" y="243" fill="#5f9fe0" font-weight="700" xml:space="preserve">█</text>
70
- <text x="172.2" y="243" fill="#e0708a" font-weight="700" xml:space="preserve">█</text>
71
- <text x="180.0" y="243" fill="#e0708a" font-weight="700" xml:space="preserve">█</text>
72
- <text x="187.8" y="243" fill="#e0708a" font-weight="700" xml:space="preserve">█</text>
73
- <text x="195.6" y="243" fill="#e0708a" font-weight="700" xml:space="preserve">█</text>
74
- <text x="203.4" y="243" fill="#d7af5f" font-weight="700" xml:space="preserve">█</text>
75
- <text x="211.2" y="243" fill="#d7af5f" font-weight="700" xml:space="preserve">█</text>
76
- <text x="219.0" y="243" fill="#d7af5f" font-weight="700" xml:space="preserve">█</text>
77
- <text x="226.8" y="243" fill="#5f9fe0" font-weight="700" xml:space="preserve">█</text>
78
- <text x="234.6" y="243" fill="#5f9fe0" font-weight="700" xml:space="preserve">█</text>
79
- <text x="242.4" y="243" fill="#5f9fe0" font-weight="700" xml:space="preserve">█</text>
80
- <text x="250.2" y="243" fill="#5f9fe0" font-weight="700" xml:space="preserve">█</text>
81
- <text x="258.0" y="243" fill="#5f9fe0" font-weight="700" xml:space="preserve">█</text>
82
- <text x="265.8" y="243" fill="#5f9fe0" font-weight="700" xml:space="preserve">█</text>
83
- <text x="273.6" y="243" fill="#5f9fe0" font-weight="700" xml:space="preserve">█</text>
84
- <text x="281.4" y="243" fill="#5f9fe0" font-weight="700" xml:space="preserve">█</text>
85
- <text x="289.2" y="243" fill="#d7af5f" font-weight="700" xml:space="preserve">█</text>
86
- <text x="297.0" y="243" fill="#d7af5f" font-weight="700" xml:space="preserve">█</text>
87
- <text x="304.8" y="243" fill="#e0708a" font-weight="700" xml:space="preserve">█</text>
88
- <text x="312.6" y="243" fill="#e0708a" font-weight="700" xml:space="preserve">█</text>
89
- <text x="320.4" y="243" fill="#e0708a" font-weight="700" xml:space="preserve">█</text>
90
- <text x="328.2" y="243" fill="#e0708a" font-weight="700" xml:space="preserve">█</text>
91
- <text x="336.0" y="243" fill="#e0708a" font-weight="700" xml:space="preserve">█</text>
92
- <text x="343.8" y="243" fill="#e0708a" font-weight="700" xml:space="preserve">█</text>
93
- <text x="351.6" y="243" fill="#e0708a" font-weight="700" xml:space="preserve">█</text>
94
- <text x="359.4" y="243" fill="#e0708a" font-weight="700" xml:space="preserve">█</text>
95
- <text x="367.2" y="243" fill="#e0708a" font-weight="700" xml:space="preserve">█</text>
96
- <text x="375.0" y="243" fill="#7c7c7c" xml:space="preserve"> 2 replies per cell</text>
97
- <text x="39.6" y="262" fill="#5f9fe0" font-weight="700" xml:space="preserve">■ </text>
98
- <text x="55.2" y="262" fill="#d0d0d0" xml:space="preserve">low 41 </text>
99
- <text x="117.6" y="262" fill="#d7af5f" font-weight="700" xml:space="preserve">■ </text>
100
- <text x="133.2" y="262" fill="#d0d0d0" xml:space="preserve">medium 12 </text>
101
- <text x="219.0" y="262" fill="#e0708a" font-weight="700" xml:space="preserve">■ </text>
102
- <text x="234.6" y="262" fill="#d0d0d0" xml:space="preserve">high 33 </text>
103
- <text x="39.6" y="281" fill="#d7875f" font-style="italic" xml:space="preserve">coding</text>
104
- <text x="86.4" y="281" fill="#7c7c7c" xml:space="preserve"> 45 · </text>
105
- <text x="133.2" y="281" fill="#d7875f" font-style="italic" xml:space="preserve">running</text>
106
- <text x="187.8" y="281" fill="#7c7c7c" xml:space="preserve"> 26 · </text>
107
- <text x="234.6" y="281" fill="#d7875f" font-style="italic" xml:space="preserve">exploring</text>
108
- <text x="304.8" y="281" fill="#7c7c7c" xml:space="preserve"> 15</text>
109
- <text x="39.6" y="300" fill="#7c7c7c" xml:space="preserve">last switch 15:38 </text>
110
- <text x="187.8" y="300" fill="#d7af5f" font-weight="700" xml:space="preserve">medium</text>
111
- <text x="234.6" y="300" fill="#7c7c7c" xml:space="preserve"> → </text>
112
- <text x="258.0" y="300" fill="#e0708a" font-weight="700" xml:space="preserve">high</text>
113
- <text x="289.2" y="300" fill="#7c7c7c" xml:space="preserve"> · </text>
114
- <text x="312.6" y="300" fill="#d7875f" font-style="italic" xml:space="preserve">coding</text>
115
- <text x="39.6" y="319" fill="#d0d0d0" xml:space="preserve">Routing saved </text>
116
- <text x="148.8" y="319" fill="#6fcf97" font-weight="700" xml:space="preserve">$4.17 (39%)</text>
117
- <text x="234.6" y="319" fill="#d0d0d0" xml:space="preserve"> vs your model</text>
118
- <text x="343.8" y="319" fill="#7c7c7c" xml:space="preserve"> · switch writes $0.19</text>
119
- <rect x="531.0" y="305" width="70.2" height="18" rx="4" fill="#2e3340"/>
120
- <text x="531.0" y="319" fill="#d0d0d0" xml:space="preserve">[ Usage ]</text>
121
- <text x="24.0" y="357" fill="#5fafd7" font-weight="700" xml:space="preserve">TODAY</text>
122
- <text x="63.0" y="357" fill="#7c7c7c" xml:space="preserve"> · 104 routed replies · 12 route switches</text>
123
- <text x="39.6" y="376" fill="#d0d0d0" xml:space="preserve">Routing saved </text>
124
- <text x="148.8" y="376" fill="#6fcf97" font-weight="700" xml:space="preserve">$5.16 (40%)</text>
125
- <text x="234.6" y="376" fill="#d0d0d0" xml:space="preserve"> vs your model</text>
126
- <text x="343.8" y="376" fill="#7c7c7c" xml:space="preserve"> · switch writes $0.23</text>
127
- <rect x="531.0" y="362" width="70.2" height="18" rx="4" fill="#2e3340"/>
128
- <text x="531.0" y="376" fill="#d0d0d0" xml:space="preserve">[ Usage ]</text>
129
- <text x="24.0" y="414" fill="#5fafd7" font-weight="700" xml:space="preserve">SUBAGENTS</text>
130
- <text x="94.2" y="414" fill="#7c7c7c" xml:space="preserve"> · this session</text>
131
- <text x="211.2" y="414" fill="#d7af5f" xml:space="preserve"> · ● 1 running</text>
132
- <rect x="336.0" y="400" width="70.2" height="18" rx="4" fill="#2e3340"/>
133
- <text x="336.0" y="414" fill="#d0d0d0" xml:space="preserve">[ Usage ]</text>
134
- <text x="39.6" y="433" fill="#c3e88d" xml:space="preserve">■■</text>
135
- <text x="55.2" y="433" fill="#c792ea" xml:space="preserve">■</text>
136
- <text x="63.0" y="433" fill="#c792ea" xml:space="preserve">□□□</text>
137
- <text x="102.0" y="433" fill="#c3e88d" xml:space="preserve">Haiku 5.5 2</text>
138
- <text x="187.8" y="433" fill="#7c7c7c" xml:space="preserve"> · </text>
139
- <text x="211.2" y="433" fill="#c792ea" xml:space="preserve">Opus 5.5 4</text>
140
- <text x="24.0" y="452" fill="#d0d0d0" xml:space="preserve"> 6 agents</text>
141
- <text x="102.0" y="452" fill="#7c7c7c" xml:space="preserve"> · </text>
142
- <text x="125.4" y="452" fill="#d0d0d0" xml:space="preserve">■ 3 moved by the router</text>
143
- <text x="304.8" y="452" fill="#7c7c7c" xml:space="preserve"> · </text>
144
- <text x="328.2" y="452" fill="#d0d0d0" xml:space="preserve">□ 3 on Claude Code’s pick</text>
145
- <text x="24.0" y="471" fill="#d0d0d0" xml:space="preserve"> Routing cost </text>
146
- <text x="141.0" y="471" fill="#f0b429" xml:space="preserve">$0.08</text>
147
- <text x="180.0" y="471" fill="#d0d0d0" xml:space="preserve"> more</text>
148
- <text x="219.0" y="471" fill="#d0d0d0" xml:space="preserve"> (7%)</text>
149
- <text x="258.0" y="471" fill="#7c7c7c" xml:space="preserve"> on 3 routed agents</text>
150
- <text x="24.0" y="509" fill="#5fafd7" font-weight="700" xml:space="preserve">WORTH A LOOK</text>
151
- <text x="117.6" y="509" fill="#7c7c7c" xml:space="preserve"> · today</text>
152
- <text x="39.6" y="528" fill="#d7875f" font-style="italic" xml:space="preserve">coding</text>
153
- <text x="86.4" y="528" fill="#7c7c7c" xml:space="preserve"> · </text>
154
- <text x="109.8" y="528" fill="#e0708a" font-weight="700" xml:space="preserve">high</text>
155
- <text x="141.0" y="528" fill="#d0d0d0" xml:space="preserve"> carried 61% of the cost</text>
156
- <text x="328.2" y="528" fill="#7c7c7c" xml:space="preserve"> · mostly Opus 5.5 · xhigh</text>
157
- <rect x="546.6" y="514" width="78.0" height="18" rx="4" fill="#2e3340"/>
158
- <text x="546.6" y="528" fill="#d0d0d0" xml:space="preserve">[ Routes ]</text>
159
- <text x="24.0" y="566" fill="#87d787" xml:space="preserve">Context ██████░░░░░░░░░░ 40% ~402.9K / 1.00M</text>
160
- <text x="24.0" y="585" fill="#87d787" xml:space="preserve">Cache </text>
161
- <text x="94.2" y="585" fill="#87d787" xml:space="preserve">100% reused on the last reply</text>
162
- <text x="320.4" y="585" fill="#7c7c7c" xml:space="preserve"> · </text>
163
- <text x="359.4" y="585" fill="#d0d0d0" xml:space="preserve">cost $18.003 </text>
164
- <text x="460.8" y="585" fill="#7c7c7c" xml:space="preserve">by Claude </text>
165
- <rect x="546.6" y="571" width="85.8" height="18" rx="4" fill="#2e3340"/>
166
- <text x="546.6" y="585" fill="#d0d0d0" xml:space="preserve">[ details ]</text>
167
- <rect x="24.0" y="590" width="70.2" height="18" rx="4" fill="#2e3340"/>
168
- <text x="24.0" y="604" fill="#d0d0d0" xml:space="preserve">[ Close ]</text>
53
+ <text x="39.6" y="243" fill="#6fcf97" font-weight="700" xml:space="preserve">routing −$4.17 (−39%)</text>
54
+ <text x="203.4" y="243" fill="#d0d0d0" xml:space="preserve"> vs Opus 5.5 · xhigh</text>
55
+ <rect x="375.0" y="229" width="70.2" height="18" rx="4" fill="#2e3340"/>
56
+ <text x="375.0" y="243" fill="#d0d0d0" xml:space="preserve">[ Usage ]</text>
57
+ <text x="55.2" y="262" fill="#d0d0d0" xml:space="preserve">Your model</text>
58
+ <text x="211.2" y="262" fill="#d0d0d0" xml:space="preserve"> $10.62</text>
59
+ <text x="304.8" y="262" fill="#e0708a" font-weight="700" xml:space="preserve">Opus 5.5 · xhigh</text>
60
+ <text x="24.0" y="281" fill="#7c7c7c" xml:space="preserve"> − </text>
61
+ <text x="55.2" y="281" fill="#d0d0d0" xml:space="preserve">Cheaper models</text>
62
+ <text x="211.2" y="281" fill="#6fcf97" xml:space="preserve"> $4.36</text>
63
+ <text x="24.0" y="300" fill="#7c7c7c" xml:space="preserve"> + </text>
64
+ <text x="55.2" y="300" fill="#d0d0d0" xml:space="preserve">Cache rebuilds</text>
65
+ <text x="211.2" y="300" fill="#f0b429" xml:space="preserve"> $0.19</text>
66
+ <text x="304.8" y="300" fill="#7c7c7c" xml:space="preserve">9 route switches · 4% of the saving</text>
67
+ <text x="24.0" y="319" fill="#7c7c7c" xml:space="preserve"> = </text>
68
+ <text x="55.2" y="319" fill="#d0d0d0" font-weight="700" xml:space="preserve">Routed</text>
69
+ <text x="211.2" y="319" fill="#d0d0d0" font-weight="700" xml:space="preserve"> $6.45</text>
70
+ <text x="304.8" y="319" fill="#5fafd7" xml:space="preserve">█████████████</text>
71
+ <text x="406.2" y="319" fill="#7c7c7c" xml:space="preserve">░░░░░░░░░</text>
72
+ <text x="476.4" y="319" fill="#7c7c7c" xml:space="preserve"> 61% of your model</text>
73
+ <text x="39.6" y="338" fill="#5f9fe0" font-weight="700" xml:space="preserve">█</text>
74
+ <text x="47.4" y="338" fill="#5f9fe0" font-weight="700" xml:space="preserve">█</text>
75
+ <text x="55.2" y="338" fill="#5f9fe0" font-weight="700" xml:space="preserve">█</text>
76
+ <text x="63.0" y="338" fill="#d7af5f" font-weight="700" xml:space="preserve">█</text>
77
+ <text x="70.8" y="338" fill="#5f9fe0" font-weight="700" xml:space="preserve">█</text>
78
+ <text x="78.6" y="338" fill="#5f9fe0" font-weight="700" xml:space="preserve">█</text>
79
+ <text x="86.4" y="338" fill="#5f9fe0" font-weight="700" xml:space="preserve">█</text>
80
+ <text x="94.2" y="338" fill="#5f9fe0" font-weight="700" xml:space="preserve">█</text>
81
+ <text x="102.0" y="338" fill="#5f9fe0" font-weight="700" xml:space="preserve">█</text>
82
+ <text x="109.8" y="338" fill="#e0708a" font-weight="700" xml:space="preserve">█</text>
83
+ <text x="117.6" y="338" fill="#e0708a" font-weight="700" xml:space="preserve">█</text>
84
+ <text x="125.4" y="338" fill="#e0708a" font-weight="700" xml:space="preserve">█</text>
85
+ <text x="133.2" y="338" fill="#e0708a" font-weight="700" xml:space="preserve">█</text>
86
+ <text x="141.0" y="338" fill="#5f9fe0" font-weight="700" xml:space="preserve">█</text>
87
+ <text x="148.8" y="338" fill="#5f9fe0" font-weight="700" xml:space="preserve">█</text>
88
+ <text x="156.6" y="338" fill="#5f9fe0" font-weight="700" xml:space="preserve">█</text>
89
+ <text x="164.4" y="338" fill="#5f9fe0" font-weight="700" xml:space="preserve">█</text>
90
+ <text x="172.2" y="338" fill="#e0708a" font-weight="700" xml:space="preserve">█</text>
91
+ <text x="180.0" y="338" fill="#e0708a" font-weight="700" xml:space="preserve">█</text>
92
+ <text x="187.8" y="338" fill="#e0708a" font-weight="700" xml:space="preserve">█</text>
93
+ <text x="195.6" y="338" fill="#e0708a" font-weight="700" xml:space="preserve">█</text>
94
+ <text x="203.4" y="338" fill="#d7af5f" font-weight="700" xml:space="preserve">█</text>
95
+ <text x="211.2" y="338" fill="#d7af5f" font-weight="700" xml:space="preserve">█</text>
96
+ <text x="219.0" y="338" fill="#d7af5f" font-weight="700" xml:space="preserve">█</text>
97
+ <text x="226.8" y="338" fill="#5f9fe0" font-weight="700" xml:space="preserve">█</text>
98
+ <text x="234.6" y="338" fill="#5f9fe0" font-weight="700" xml:space="preserve">█</text>
99
+ <text x="242.4" y="338" fill="#5f9fe0" font-weight="700" xml:space="preserve">█</text>
100
+ <text x="250.2" y="338" fill="#5f9fe0" font-weight="700" xml:space="preserve">█</text>
101
+ <text x="258.0" y="338" fill="#5f9fe0" font-weight="700" xml:space="preserve">█</text>
102
+ <text x="265.8" y="338" fill="#5f9fe0" font-weight="700" xml:space="preserve">█</text>
103
+ <text x="273.6" y="338" fill="#5f9fe0" font-weight="700" xml:space="preserve">█</text>
104
+ <text x="281.4" y="338" fill="#5f9fe0" font-weight="700" xml:space="preserve">█</text>
105
+ <text x="289.2" y="338" fill="#d7af5f" font-weight="700" xml:space="preserve">█</text>
106
+ <text x="297.0" y="338" fill="#d7af5f" font-weight="700" xml:space="preserve">█</text>
107
+ <text x="304.8" y="338" fill="#e0708a" font-weight="700" xml:space="preserve">█</text>
108
+ <text x="312.6" y="338" fill="#e0708a" font-weight="700" xml:space="preserve">█</text>
109
+ <text x="320.4" y="338" fill="#e0708a" font-weight="700" xml:space="preserve">█</text>
110
+ <text x="328.2" y="338" fill="#e0708a" font-weight="700" xml:space="preserve">█</text>
111
+ <text x="336.0" y="338" fill="#e0708a" font-weight="700" xml:space="preserve">█</text>
112
+ <text x="343.8" y="338" fill="#e0708a" font-weight="700" xml:space="preserve">█</text>
113
+ <text x="351.6" y="338" fill="#e0708a" font-weight="700" xml:space="preserve">█</text>
114
+ <text x="359.4" y="338" fill="#e0708a" font-weight="700" xml:space="preserve">█</text>
115
+ <text x="367.2" y="338" fill="#e0708a" font-weight="700" xml:space="preserve">█</text>
116
+ <text x="375.0" y="338" fill="#7c7c7c" xml:space="preserve"> 2 replies per cell</text>
117
+ <text x="39.6" y="357" fill="#5f9fe0" font-weight="700" xml:space="preserve">■ </text>
118
+ <text x="55.2" y="357" fill="#d0d0d0" xml:space="preserve">low 41 </text>
119
+ <text x="117.6" y="357" fill="#d7af5f" font-weight="700" xml:space="preserve">■ </text>
120
+ <text x="133.2" y="357" fill="#d0d0d0" xml:space="preserve">medium 12 </text>
121
+ <text x="219.0" y="357" fill="#e0708a" font-weight="700" xml:space="preserve">■ </text>
122
+ <text x="234.6" y="357" fill="#d0d0d0" xml:space="preserve">high 33 </text>
123
+ <text x="39.6" y="376" fill="#d7875f" font-style="italic" xml:space="preserve">coding</text>
124
+ <text x="86.4" y="376" fill="#7c7c7c" xml:space="preserve"> 45 · </text>
125
+ <text x="133.2" y="376" fill="#d7875f" font-style="italic" xml:space="preserve">running</text>
126
+ <text x="187.8" y="376" fill="#7c7c7c" xml:space="preserve"> 26 · </text>
127
+ <text x="234.6" y="376" fill="#d7875f" font-style="italic" xml:space="preserve">exploring</text>
128
+ <text x="304.8" y="376" fill="#7c7c7c" xml:space="preserve"> 15</text>
129
+ <text x="39.6" y="395" fill="#7c7c7c" xml:space="preserve">last switch 15:38 </text>
130
+ <text x="187.8" y="395" fill="#d7af5f" font-weight="700" xml:space="preserve">medium</text>
131
+ <text x="234.6" y="395" fill="#7c7c7c" xml:space="preserve"> → </text>
132
+ <text x="258.0" y="395" fill="#e0708a" font-weight="700" xml:space="preserve">high</text>
133
+ <text x="289.2" y="395" fill="#7c7c7c" xml:space="preserve"> · </text>
134
+ <text x="312.6" y="395" fill="#d7875f" font-style="italic" xml:space="preserve">coding</text>
135
+ <text x="24.0" y="433" fill="#5fafd7" font-weight="700" xml:space="preserve">TODAY</text>
136
+ <text x="63.0" y="433" fill="#7c7c7c" xml:space="preserve"> · 104 routed replies · 12 route switches</text>
137
+ <text x="39.6" y="452" fill="#6fcf97" font-weight="700" xml:space="preserve">routing −$5.16 (−40%)</text>
138
+ <text x="203.4" y="452" fill="#d0d0d0" xml:space="preserve"> vs each session’s own model</text>
139
+ <rect x="437.4" y="438" width="70.2" height="18" rx="4" fill="#2e3340"/>
140
+ <text x="437.4" y="452" fill="#d0d0d0" xml:space="preserve">[ Usage ]</text>
141
+ <text x="24.0" y="490" fill="#5fafd7" font-weight="700" xml:space="preserve">SUBAGENTS</text>
142
+ <text x="94.2" y="490" fill="#7c7c7c" xml:space="preserve"> · this session</text>
143
+ <text x="211.2" y="490" fill="#d7af5f" xml:space="preserve"> · ● 1 running</text>
144
+ <rect x="336.0" y="476" width="70.2" height="18" rx="4" fill="#2e3340"/>
145
+ <text x="336.0" y="490" fill="#d0d0d0" xml:space="preserve">[ Usage ]</text>
146
+ <text x="39.6" y="509" fill="#c3e88d" xml:space="preserve">■■</text>
147
+ <text x="55.2" y="509" fill="#c792ea" xml:space="preserve">■</text>
148
+ <text x="63.0" y="509" fill="#c792ea" xml:space="preserve">□□□</text>
149
+ <text x="102.0" y="509" fill="#c3e88d" xml:space="preserve">Haiku 5.5 2</text>
150
+ <text x="187.8" y="509" fill="#7c7c7c" xml:space="preserve"> · </text>
151
+ <text x="211.2" y="509" fill="#c792ea" xml:space="preserve">Opus 5.5 4</text>
152
+ <text x="24.0" y="528" fill="#d0d0d0" xml:space="preserve"> 6 agents</text>
153
+ <text x="102.0" y="528" fill="#7c7c7c" xml:space="preserve"> · </text>
154
+ <text x="125.4" y="528" fill="#d0d0d0" xml:space="preserve">■ 3 moved by the router</text>
155
+ <text x="304.8" y="528" fill="#7c7c7c" xml:space="preserve"> · </text>
156
+ <text x="328.2" y="528" fill="#d0d0d0" xml:space="preserve">□ 3 on Claude Code’s pick</text>
157
+ <text x="39.6" y="547" fill="#f0b429" font-weight="700" xml:space="preserve">routing +$0.08 (+7%)</text>
158
+ <text x="195.6" y="547" fill="#d0d0d0" xml:space="preserve"> vs Claude Code’s picks</text>
159
+ <text x="375.0" y="547" fill="#7c7c7c" xml:space="preserve"> on 3 routed agents</text>
160
+ <text x="24.0" y="585" fill="#5fafd7" font-weight="700" xml:space="preserve">WORTH A LOOK</text>
161
+ <text x="117.6" y="585" fill="#7c7c7c" xml:space="preserve"> · today</text>
162
+ <text x="39.6" y="604" fill="#d7875f" font-style="italic" xml:space="preserve">coding</text>
163
+ <text x="86.4" y="604" fill="#7c7c7c" xml:space="preserve"> · </text>
164
+ <text x="109.8" y="604" fill="#e0708a" font-weight="700" xml:space="preserve">high</text>
165
+ <text x="141.0" y="604" fill="#d0d0d0" xml:space="preserve"> carried 61% of the cost</text>
166
+ <text x="328.2" y="604" fill="#7c7c7c" xml:space="preserve"> · mostly Opus 5.5 · xhigh</text>
167
+ <rect x="546.6" y="590" width="78.0" height="18" rx="4" fill="#2e3340"/>
168
+ <text x="546.6" y="604" fill="#d0d0d0" xml:space="preserve">[ Routes ]</text>
169
+ <text x="24.0" y="642" fill="#87d787" xml:space="preserve">Context ██████░░░░░░░░░░ 40% ~402.9K / 1.00M</text>
170
+ <text x="24.0" y="661" fill="#87d787" xml:space="preserve">Cache </text>
171
+ <text x="94.2" y="661" fill="#87d787" xml:space="preserve">100% reused on the last reply</text>
172
+ <text x="320.4" y="661" fill="#7c7c7c" xml:space="preserve"> · </text>
173
+ <text x="359.4" y="661" fill="#d0d0d0" xml:space="preserve">cost $18.003 </text>
174
+ <text x="460.8" y="661" fill="#7c7c7c" xml:space="preserve">by Claude </text>
175
+ <rect x="546.6" y="647" width="85.8" height="18" rx="4" fill="#2e3340"/>
176
+ <text x="546.6" y="661" fill="#d0d0d0" xml:space="preserve">[ details ]</text>
177
+ <rect x="24.0" y="666" width="70.2" height="18" rx="4" fill="#2e3340"/>
178
+ <text x="24.0" y="680" fill="#d0d0d0" xml:space="preserve">[ Close ]</text>
169
179
  </svg>
@@ -87,7 +87,7 @@
87
87
  <text x="24.0" y="452" fill="#7c7c7c" xml:space="preserve">cold cache writes the whole context again, the switch cost on Now.</text>
88
88
  <rect x="24.0" y="457" width="218.4" height="18" rx="4" fill="#2e3340"/>
89
89
  <text x="24.0" y="471" fill="#d0d0d0" xml:space="preserve">[ Reset routes to defaults ]</text>
90
- <text x="24.0" y="490" fill="#7c7c7c" xml:space="preserve">A new model ID needs router.json → models: price, window, efforts.</text>
90
+ <text x="24.0" y="490" fill="#7c7c7c" xml:space="preserve">A new model ID needs the active file → models: price, window, efforts.</text>
91
91
  <text x="24.0" y="528" fill="#7c7c7c" xml:space="preserve">router.json changes:</text>
92
92
  <text x="24.0" y="547" fill="#e06c6c" xml:space="preserve">- routes.medium Opus 5.5 · medium</text>
93
93
  <text x="24.0" y="566" fill="#87d787" xml:space="preserve">+ routes.medium Sonnet 5.5 · xhigh</text>
@@ -40,9 +40,8 @@
40
40
  <text x="453.0" y="91" fill="#7c7c7c" xml:space="preserve"> p next</text>
41
41
  <text x="24.0" y="110" fill="#7c7c7c" xml:space="preserve">this session · your model </text>
42
42
  <text x="226.8" y="110" fill="#e0708a" font-weight="700" xml:space="preserve">Opus 5.5 · xhigh</text>
43
- <text x="24.0" y="148" fill="#d0d0d0" xml:space="preserve">Routing saved </text>
44
- <text x="133.2" y="148" fill="#6fcf97" font-weight="700" xml:space="preserve">$4.17 (39%)</text>
45
- <text x="219.0" y="148" fill="#d0d0d0" xml:space="preserve"> vs your model</text>
43
+ <text x="24.0" y="148" fill="#6fcf97" font-weight="700" xml:space="preserve">routing −$4.17 (−39%)</text>
44
+ <text x="187.8" y="148" fill="#d0d0d0" xml:space="preserve"> vs Opus 5.5 · xhigh</text>
46
45
  <text x="55.2" y="186" fill="#d0d0d0" xml:space="preserve">Your model</text>
47
46
  <text x="211.2" y="186" fill="#d0d0d0" xml:space="preserve"> $10.62</text>
48
47
  <text x="304.8" y="186" fill="#e0708a" font-weight="700" xml:space="preserve">Opus 5.5 · xhigh</text>
@@ -53,7 +52,7 @@
53
52
  <text x="55.2" y="224" fill="#d0d0d0" xml:space="preserve">Stronger models</text>
54
53
  <text x="211.2" y="224" fill="#d0d0d0" xml:space="preserve"> $0.00</text>
55
54
  <text x="24.0" y="243" fill="#7c7c7c" xml:space="preserve"> + </text>
56
- <text x="55.2" y="243" fill="#d0d0d0" xml:space="preserve">Switch cache writes</text>
55
+ <text x="55.2" y="243" fill="#d0d0d0" xml:space="preserve">Cache rebuilds</text>
57
56
  <text x="211.2" y="243" fill="#f0b429" xml:space="preserve"> $0.19</text>
58
57
  <text x="304.8" y="243" fill="#7c7c7c" xml:space="preserve">9 route switches · 4% of the saving</text>
59
58
  <text x="24.0" y="262" fill="#7c7c7c" xml:space="preserve"> = </text>
@@ -133,10 +132,9 @@
133
132
  <text x="133.2" y="680" fill="#d0d0d0" xml:space="preserve">■ 8 moved by the router</text>
134
133
  <text x="312.6" y="680" fill="#7c7c7c" xml:space="preserve"> · </text>
135
134
  <text x="336.0" y="680" fill="#d0d0d0" xml:space="preserve">□ 7 on Claude Code’s pick</text>
136
- <text x="24.0" y="699" fill="#d0d0d0" xml:space="preserve"> Routing saved </text>
137
- <text x="148.8" y="699" fill="#6fcf97" xml:space="preserve">$0.51</text>
138
- <text x="187.8" y="699" fill="#d0d0d0" xml:space="preserve"> (14%)</text>
139
- <text x="234.6" y="699" fill="#7c7c7c" xml:space="preserve"> on 9 routed agents</text>
135
+ <text x="39.6" y="699" fill="#6fcf97" font-weight="700" xml:space="preserve">routing −$0.51 (−14%)</text>
136
+ <text x="203.4" y="699" fill="#d0d0d0" xml:space="preserve"> vs Claude Code’s picks</text>
137
+ <text x="382.8" y="699" fill="#7c7c7c" xml:space="preserve"> on 9 routed agents</text>
140
138
  <text x="24.0" y="737" fill="#d0d0d0" font-weight="700" xml:space="preserve"> ROUTED</text>
141
139
  <text x="86.4" y="737" fill="#7c7c7c" xml:space="preserve"> · the router picked the model at spawn</text>
142
140
  <text x="24.0" y="756" fill="#7c7c7c" xml:space="preserve"> type agents cost vs CC</text>
@@ -39,9 +39,8 @@
39
39
  <text x="390.6" y="91" fill="#d0d0d0" xml:space="preserve">[ All ]</text>
40
40
  <text x="453.0" y="91" fill="#7c7c7c" xml:space="preserve"> p next</text>
41
41
  <text x="24.0" y="110" fill="#7c7c7c" xml:space="preserve">all projects · Oct 7 – Oct 13 · each session vs its own model</text>
42
- <text x="24.0" y="148" fill="#d0d0d0" xml:space="preserve">Routing saved </text>
43
- <text x="133.2" y="148" fill="#6fcf97" font-weight="700" xml:space="preserve">$11.78 (43%)</text>
44
- <text x="226.8" y="148" fill="#d0d0d0" xml:space="preserve"> vs your model</text>
42
+ <text x="24.0" y="148" fill="#6fcf97" font-weight="700" xml:space="preserve">routing −$11.78 (−43%)</text>
43
+ <text x="195.6" y="148" fill="#d0d0d0" xml:space="preserve"> vs each session’s own model</text>
45
44
  <text x="55.2" y="186" fill="#d0d0d0" xml:space="preserve">Your model</text>
46
45
  <text x="211.2" y="186" fill="#d0d0d0" xml:space="preserve"> $27.11</text>
47
46
  <text x="304.8" y="186" fill="#7c7c7c" xml:space="preserve">each session’s own model</text>
@@ -52,7 +51,7 @@
52
51
  <text x="55.2" y="224" fill="#d0d0d0" xml:space="preserve">Stronger models</text>
53
52
  <text x="211.2" y="224" fill="#d0d0d0" xml:space="preserve"> $0.00</text>
54
53
  <text x="24.0" y="243" fill="#7c7c7c" xml:space="preserve"> + </text>
55
- <text x="55.2" y="243" fill="#d0d0d0" xml:space="preserve">Switch cache writes</text>
54
+ <text x="55.2" y="243" fill="#d0d0d0" xml:space="preserve">Cache rebuilds</text>
56
55
  <text x="211.2" y="243" fill="#f0b429" xml:space="preserve"> $0.51</text>
57
56
  <text x="304.8" y="243" fill="#7c7c7c" xml:space="preserve">30 route switches · 4% of the saving</text>
58
57
  <text x="24.0" y="262" fill="#7c7c7c" xml:space="preserve"> = </text>
@@ -129,10 +128,9 @@
129
128
  <text x="133.2" y="661" fill="#d0d0d0" xml:space="preserve">■ 8 moved by the router</text>
130
129
  <text x="312.6" y="661" fill="#7c7c7c" xml:space="preserve"> · </text>
131
130
  <text x="336.0" y="661" fill="#d0d0d0" xml:space="preserve">□ 7 on Claude Code’s pick</text>
132
- <text x="24.0" y="680" fill="#d0d0d0" xml:space="preserve"> Routing saved </text>
133
- <text x="148.8" y="680" fill="#6fcf97" xml:space="preserve">$0.51</text>
134
- <text x="187.8" y="680" fill="#d0d0d0" xml:space="preserve"> (14%)</text>
135
- <text x="234.6" y="680" fill="#7c7c7c" xml:space="preserve"> on 9 routed agents</text>
131
+ <text x="39.6" y="680" fill="#6fcf97" font-weight="700" xml:space="preserve">routing −$0.51 (−14%)</text>
132
+ <text x="203.4" y="680" fill="#d0d0d0" xml:space="preserve"> vs Claude Code’s picks</text>
133
+ <text x="382.8" y="680" fill="#7c7c7c" xml:space="preserve"> on 9 routed agents</text>
136
134
  <text x="24.0" y="699" fill="#d0d0d0" xml:space="preserve"> per day </text>
137
135
  <text x="102.0" y="699" fill="#d0d0d0" font-weight="700" xml:space="preserve">▁▂▂▃▃▅█</text>
138
136
  <text x="156.6" y="699" fill="#7c7c7c" xml:space="preserve"> peak 6 on Oct 13</text>