From c2d90db1e31165095b6f7fa770d4464cc1b52100 Mon Sep 17 00:00:00 2001
From: Mohammad Miadh Angkad <176301910+mmangkad@users.noreply.github.com>
Date: Sun, 9 Aug 2026 04:42:59 +0800
Subject: [PATCH] [CI] Add Kimi-K3 low-latency performance check (#34089)

---
 test/registered/models_e2e/test_kimi_k3_b300.py | 8 ++++++--
 1 file changed, 6 insertions(+), 2 deletions(-)

diff --git a/test/registered/models_e2e/test_kimi_k3_b300.py b/test/registered/models_e2e/test_kimi_k3_b300.py
index 9a7a378c1..b38004670 100644
--- a/test/registered/models_e2e/test_kimi_k3_b300.py
+++ b/test/registered/models_e2e/test_kimi_k3_b300.py
@@ -1,7 +1,8 @@
 """B300 per-commit CI coverage for Kimi-K3 serving recipes.
 
 Runs the Low Latency DSPARK recipe and the Balanced DCP/HiCache recipe on
-eight B300 GPUs. Each server must preserve basic model quality on GSM8K.
+eight B300 GPUs. Each server must preserve basic model quality on GSM8K, and
+the Low Latency recipe must also preserve single-request decode performance.
 """
 
 import unittest
@@ -9,6 +10,7 @@ import unittest
 from sglang.srt.utils import kill_process_tree
 from sglang.test.ci.ci_register import register_cuda_ci
 from sglang.test.kits.eval_accuracy_kit import GSM8KMixin
+from sglang.test.kits.spec_decoding_kit import SpecDecodingMixin
 from sglang.test.test_utils import (
     DEFAULT_URL_FOR_TEST,
     CustomTestCase,
@@ -33,11 +35,13 @@ def _stop_server(process):
         _wait_for_gpu_idle_in_ci(timeout=GPU_IDLE_TIMEOUT)
 
 
-class TestKimiK3B300LowLatency(GSM8KMixin, CustomTestCase):
+class TestKimiK3B300LowLatency(GSM8KMixin, SpecDecodingMixin, CustomTestCase):
     """TP8 Low Latency recipe with DSPARK linear ReplaySSM speculation."""
 
     gsm8k_score_threshold = 0.95
     gsm8k_num_examples = 200
+    accept_length_thres = 6.6
+    bs_1_speed_thres = 440
 
     @classmethod
     def setUpClass(cls):