You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
Commit 246f14e
Browse filesBrowse the repository at this point in the historyBrowse files
Copy file name to clipboardExpand all lines: results/reports/all/README.md
+19-19Lines changed: 19 additions & 19 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -18,7 +18,7 @@ Every number below is regenerated from the raw `results/*.json` - re-run `run-al
18
18
|[Portkey](https://github.com/Portkey-AI/gateway)| 3,476 µs | 877 <sub>(+0% from 5 ms to no bound)</sub> | 124 MiB | 244 MiB |`portkeyai/gateway:1.15.2`|
**≥** = the sweep's top rung won, so that rate is a **floor** and no ceiling was established. **0** = the sweep ran and no rung held that bound while failing nothing. **n/a** = the record carries no reading at that bound. A **✕** cell names the record's own reason for the absence.
50
50
@@ -66,7 +66,7 @@ Every rung of the same sweep the frontier readings above are taken from, summari
66
66
|[Portkey](https://github.com/Portkey-AI/gateway)| 861 at c=1 | 879 at c=8 | 1.0× / 8× | c=1 | 3.41 ms → 2.18 s | c=1024 | c=16384 |
67
67
|[TensorZero](https://github.com/tensorzero/tensorzero)| 24 at c=1 | 13,747 at c=2048 | 572.8× / 2048× | c=1024 | 41 ms → 55 ms | c=1024 | c=2048 |
68
68
|[Plano](https://github.com/katanemo/plano)| 4 at c=1 | 21 at c=8 | 5.2× / 8× | c=8 | 199 ms → 5.68 s | none | c=128 |
69
-
|[One-API](https://github.com/songquanpeng/one-api)| 29 at c=1 |40 at c=32 | 1.4× / 32× | c=32 | 42.4 ms → 3.26 s | c=4| c=32 |
69
+
|[One-API](https://github.com/songquanpeng/one-api)| 29 at c=1 |42 at c=32 | 1.4× / 32× | c=32 | 42.4 ms → 3.18 s | c=16| c=32 |
70
70
71
71
A rung that failed a request it had accepted qualifies for **no** frontier reading at any bound, so rate measured at or above the failing concurrency is not throughput the board will publish - the climb chart rules that region off. **none** in that column is a measured result across the whole ladder, not a missing one.
72
72
@@ -88,39 +88,39 @@ Same box, same mock, one gateway at a time. Streaming figures are the overhead t
**✕** cells are measured refusals, not gaps: the gateway was offered the load and could not do the thing (buffered instead of streaming, rejected the Anthropic shape, or has no native key/limit governance). **n/a** = that suite hasn't been run for this gateway yet.
Method: added latency = gateway p99 − direct-to-mock p99 at concurrency 1; a frontier reading = the highest req/s any probed concurrency carried while 99% of requests finished under the STATED bound and the gateway failed none it accepted (readings are published at 1, 5, 10, 50, 100 ms and with no bound; the columns above use 10 ms, and every caption names the bound it used); cost figures divide that 10 ms reading by $0.1632/hr for the pinned 4-core (m7g.xlarge) slice; RSS idle = after first 200, steady state = the level the RSS settled at under load. Same box, same mock, same load, one gateway at a time. Each gateway's source ref is pinned in its own `gateways/<name>/definition.json`; the built commit is in each row.
125
125
126
-
<sub>Page + charts regenerated **2026-07-30 18:59 UTC** from the raw `results/*.json`.</sub>
126
+
<sub>Page + charts regenerated **2026-07-30 20:13 UTC** from the raw `results/*.json`.</sub>
Copy file name to clipboardExpand all lines: results/reports/top5/README.md
+12-12Lines changed: 12 additions & 12 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -55,29 +55,29 @@ Same box, same mock, one gateway at a time. Streaming figures are the overhead t
55
55
56
56
**✕** cells are measured refusals, not gaps: the gateway was offered the load and could not do the thing (buffered instead of streaming, rejected the Anthropic shape, or has no native key/limit governance). **n/a** = that suite hasn't been run for this gateway yet.
Method: added latency = gateway p99 − direct-to-mock p99 at concurrency 1; a frontier reading = the highest req/s any probed concurrency carried while 99% of requests finished under the STATED bound and the gateway failed none it accepted (readings are published at 1, 5, 10, 50, 100 ms and with no bound; the columns above use 10 ms, and every caption names the bound it used); cost figures divide that 10 ms reading by $0.1632/hr for the pinned 4-core (m7g.xlarge) slice; RSS idle = after first 200, steady state = the level the RSS settled at under load. Same box, same mock, same load, one gateway at a time. Each gateway's source ref is pinned in its own `gateways/<name>/definition.json`; the built commit is in each row.
82
82
83
-
<sub>Page + charts regenerated **2026-07-30 18:59 UTC** from the raw `results/*.json`.</sub>
83
+
<sub>Page + charts regenerated **2026-07-30 20:13 UTC** from the raw `results/*.json`.</sub>
![github-actions[bot]](https://avatars.githubusercontent.com/in/15368?v=4&size=40){kind=avatar}
0 commit comments