From c2e8da1e65d92392c7e27de9faf67890bacf4173 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Michael=20M=C3=B6bius?= <e-mail@rocketrider.eu>
Date: Sat, 4 Apr 2026 09:19:51 +0200
Subject: [PATCH] Add truncate_prompt_tokens param for vllm embedding

vllm embedding should allow sending "truncate_prompt_tokens". Links to https://github.com/vllm-project/vllm/blob/main/vllm/entrypoints/pooling/embed/protocol.py#L66
---
 litellm/llms/hosted_vllm/embedding/transformation.py | 1 +
 1 file changed, 1 insertion(+)

diff --git a/litellm/llms/hosted_vllm/embedding/transformation.py b/litellm/llms/hosted_vllm/embedding/transformation.py
index 9c3e8c6c7cc..0db3e820a85 100644
--- a/litellm/llms/hosted_vllm/embedding/transformation.py
+++ b/litellm/llms/hosted_vllm/embedding/transformation.py
@@ -150,6 +150,7 @@ class HostedVLLMEmbeddingConfig(BaseEmbeddingConfig):
             "dimensions",
             "encoding_format",
             "user",
+            "truncate_prompt_tokens",
         ]
 
     def map_openai_params(