From 3d72944fb89842672cf9d663228663e05096d8c6 Mon Sep 17 00:00:00 2001
From: b8zhong <b8zhong@uwaterloo.ca>
Date: Thu, 15 Jan 2026 13:30:18 -0800
Subject: [PATCH] [Doc] Add tip on how to use Spec V2 (#15455)

---
 docs/basic_usage/deepseek_v3.md |  4 ++++
 docs/basic_usage/glm45.md       | 12 +++++-------
 docs/basic_usage/gpt_oss.md     | 18 ++++++++++++++++++
 3 files changed, 27 insertions(+), 7 deletions(-)

diff --git a/docs/basic_usage/deepseek_v3.md b/docs/basic_usage/deepseek_v3.md
index 1e406ee90..a321eb09c 100644
--- a/docs/basic_usage/deepseek_v3.md
+++ b/docs/basic_usage/deepseek_v3.md
@@ -193,6 +193,10 @@ To enable DeepSeek MTP for large batch sizes (>48), you need to adjust some para
 - Set `--cuda-graph-bs`. It's a list of batch sizes for cuda graph capture. The [default captured batch sizes for speculative decoding](https://github.com/sgl-project/sglang/blob/main/python/sglang/srt/server_args.py#L888-L895) is 48. You can customize this by including more batch sizes.
 ```
 
+```{tip}
+To enable the experimental overlap scheduler for EAGLE speculative decoding, set the environment variable `SGLANG_ENABLE_SPEC_V2=1`. This can improve performance by enabling overlap scheduling between draft and verification stages.
+```
+
 
 ### Reasoning Content for DeepSeek R1 & V3.1
 
diff --git a/docs/basic_usage/glm45.md b/docs/basic_usage/glm45.md
index b68f984f9..aaf8e13b4 100644
--- a/docs/basic_usage/glm45.md
+++ b/docs/basic_usage/glm45.md
@@ -6,11 +6,6 @@ To serve GLM-4.5 / GLM-4.6 FP8 models on 8xH100/H200 GPUs:
 python3 -m sglang.launch_server --model zai-org/GLM-4.6-FP8 --tp 8
 ```
 
-### Configuration Tips
-
-- `--max-mamba-cache-size`: Adjust `--max-mamba-cache-size` to increase mamba cache space and max running requests
-  capability. It will decrease KV cache space as a trade-off. You can adjust it according to workload.
-
 ### EAGLE Speculative Decoding
 
 **Description**: SGLang has supported GLM-4.5 / GLM-4.6 models
@@ -35,9 +30,12 @@ python3 -m sglang.launch_server \
   --enable-custom-logit-processor
 ```
 
-**Note**: For GLM-4.7, `--tool-call-parser` should be set to `glm47`, for GLM-4.5 and GLM-4.6, it should be set to `glm45`.
+```{tip}
+To enable the experimental overlap scheduler for EAGLE speculative decoding, set the environment variable `SGLANG_ENABLE_SPEC_V2=1`. This can improve performance by enabling overlap scheduling between draft and verification stages.
+```
 
-### Thinking Budget
+### Thinking Budget for GLM-4.5 / GLM-4.6
+**Note**: For GLM-4.7, `--tool-call-parser` should be set to `glm47`, for GLM-4.5 and GLM-4.6, it should be set to `glm45`.
 
 In SGLang, we can implement thinking budget with `CustomLogitProcessor`.
 
diff --git a/docs/basic_usage/gpt_oss.md b/docs/basic_usage/gpt_oss.md
index d1af32f5f..f74ba40d9 100644
--- a/docs/basic_usage/gpt_oss.md
+++ b/docs/basic_usage/gpt_oss.md
@@ -59,6 +59,24 @@ python -m sglang.launch_server ... --tool-server ip-1:port-1,ip-2:port-2
 ```
 The URLs should be MCP SSE servers that expose server information and well-documented tools. These tools are added to the system prompt so the model can use them.
 
+## Speculative Decoding
+
+SGLang supports speculative decoding for GPT-OSS models using EAGLE3 algorithm. This can significantly improve decoding speed, especially for small batch sizes.
+
+**Usage**:
+Add `--speculative-algorithm EAGLE3` along with the draft model path.
+```bash
+python3 -m sglang.launch_server \
+  --model-path openai/gpt-oss-120b \
+  --speculative-algorithm EAGLE3 \
+  --speculative-draft-model-path lmsys/EAGLE3-gpt-oss-120b-bf16 \
+  --tp 2
+```
+
+```{tip}
+To enable the experimental overlap scheduler for EAGLE3 speculative decoding, set the environment variable `SGLANG_ENABLE_SPEC_V2=1`. This can improve performance by enabling overlap scheduling between draft and verification stages.
+```
+
 ### Quick Demo
 
 ```python