Add LLM Gateway FAQ section with LeMUR migration article

fern/docs.yml

@@ -1263,6 +1263,12 @@ navigation:
               - page: Understanding Input and Output Tokens for LLM Gateway
                 path: pages/faq/speech-understanding/understanding-input-and-output-tokens-in-ai-models.mdx
                 slug: /understanding-input-and-output-tokens-in-ai-models
+          - section: LLM Gateway
+            skip-slug: true
+            contents:
+              - page: How do I switch from using LeMUR to using LLM Gateway?
+                path: pages/faq/llm-gateway/how-do-i-switch-from-lemur-to-llm-gateway.mdx
+                slug: /how-do-i-switch-from-lemur-to-llm-gateway
           - section: Playground
             skip-slug: true
             contents:

fern/pages/faq/llm-gateway/how-do-i-switch-from-lemur-to-llm-gateway.mdx

@@ -0,0 +1,9 @@
+---
+title: "How do I switch from using LeMUR to using LLM Gateway?"
+---
+
+To switch from LeMUR to LLM Gateway, you need to make three key changes: include the transcript text directly in your request instead of passing transcript IDs, update the model name format, and modify your response handling to use the OpenAI-compatible format.
+
+With LeMUR, you passed `transcript_ids` and the system automatically fetched the transcript content. With LLM Gateway, you first retrieve the transcript text and include it in your message content when calling the `/v1/chat/completions` endpoint at `llm-gateway.assemblyai.com`.
+
+For a detailed step-by-step walkthrough with code examples in Python and JavaScript, see the [Migration Guide: From LeMUR to LLM Gateway](/docs/llm-gateway/migration-from-lemur).