From c4e30593b16fa57ebe7230f2181b35149a3689c6 Mon Sep 17 00:00:00 2001
From: Anant Goel <anant@zed.dev>
Date: Wed, 17 Jun 2026 10:35:47 -0700
Subject: [PATCH] language_models: Clear speed for OpenAI-compatible providers
 that don't support fast mode (#59496)

`OpenAiCompatibleLanguageModel::stream_completion` (used for custom
OpenAI-compatible providers) forwarded `request.speed` straight into
`into_open_ai`/`into_open_ai_response`, which translate `Speed::Fast`
into OpenAI's `service_tier` field. Custom OpenAI-compatible endpoints
don't recognize that field and reject the request.

`speed` can end up set even for a provider that doesn't support fast
mode: `Thread::inherit_parent_settings` copies `speed` from a parent
thread to a subagent without checking whether the subagent's model
supports it.

Every other provider sharing this conversion code already guards against
this (`open_ai.rs`, `anthropic.rs`, `anthropic_compatible.rs`,
`language_models_cloud.rs`). This adds the same guard for
`OpenAiCompatibleLanguageModel`.

Release Notes:

- Fixed requests to OpenAI-compatible providers (e.g. Baseten) sometimes
including an unsupported `service_tier` parameter
---
 crates/language_models/src/provider/open_ai_compatible.rs | 8 +++++++-
 1 file changed, 7 insertions(+), 1 deletion(-)

diff --git a/crates/language_models/src/provider/open_ai_compatible.rs b/crates/language_models/src/provider/open_ai_compatible.rs
index c1f9f70a15..da165a77b5 100644
--- a/crates/language_models/src/provider/open_ai_compatible.rs
+++ b/crates/language_models/src/provider/open_ai_compatible.rs
@@ -322,7 +322,7 @@ impl LanguageModel for OpenAiCompatibleLanguageModel {
 
     fn stream_completion(
         &self,
-        request: LanguageModelRequest,
+        mut request: LanguageModelRequest,
         cx: &AsyncApp,
     ) -> BoxFuture<
         'static,
@@ -334,6 +334,12 @@ impl LanguageModel for OpenAiCompatibleLanguageModel {
             LanguageModelCompletionError,
         >,
     > {
+        // `speed` can leak in from a parent thread's model; this provider never
+        // supports fast mode, and arbitrary compatible endpoints reject `service_tier`.
+        if !self.supports_fast_mode() {
+            request.speed = None;
+        }
+
         if self.model.capabilities.chat_completions {
             let request = into_open_ai(
                 request,