diff --git a/crates/cloud_llm_client/src/cloud_llm_client.rs b/crates/cloud_llm_client/src/cloud_llm_client.rs
index b4429c0aa0..167507e86a 100644
--- a/crates/cloud_llm_client/src/cloud_llm_client.rs
+++ b/crates/cloud_llm_client/src/cloud_llm_client.rs
@@ -91,7 +91,6 @@ impl FromStr for UsageLimit {
 #[strum(serialize_all = "snake_case")]
 pub enum LanguageModelProvider {
     Anthropic,
-    Baseten,
     OpenAi,
     Google,
     XAi,
diff --git a/crates/language_model_core/src/provider.rs b/crates/language_model_core/src/provider.rs
index b32d2592c8..317a0ab280 100644
--- a/crates/language_model_core/src/provider.rs
+++ b/crates/language_model_core/src/provider.rs
@@ -5,10 +5,6 @@ pub const ANTHROPIC_PROVIDER_ID: LanguageModelProviderId =
 pub const ANTHROPIC_PROVIDER_NAME: LanguageModelProviderName =
     LanguageModelProviderName::new("Anthropic");
 
-pub const BASETEN_PROVIDER_ID: LanguageModelProviderId = LanguageModelProviderId::new("baseten");
-pub const BASETEN_PROVIDER_NAME: LanguageModelProviderName =
-    LanguageModelProviderName::new("Baseten");
-
 pub const OPEN_AI_PROVIDER_ID: LanguageModelProviderId = LanguageModelProviderId::new("openai");
 pub const OPEN_AI_PROVIDER_NAME: LanguageModelProviderName =
     LanguageModelProviderName::new("OpenAI");
@@ -29,8 +25,6 @@ pub fn provider_name_for_id(provider_id: &LanguageModelProviderId) -> LanguageMo
         OPEN_AI_PROVIDER_NAME
     } else if provider_id == &ANTHROPIC_PROVIDER_ID {
         ANTHROPIC_PROVIDER_NAME
-    } else if provider_id == &BASETEN_PROVIDER_ID {
-        BASETEN_PROVIDER_NAME
     } else if provider_id == &GOOGLE_PROVIDER_ID {
         GOOGLE_PROVIDER_NAME
     } else if provider_id == &X_AI_PROVIDER_ID {
diff --git a/crates/language_models_cloud/src/language_models_cloud.rs b/crates/language_models_cloud/src/language_models_cloud.rs
index a98dfb00df..a3dbd57cd6 100644
--- a/crates/language_models_cloud/src/language_models_cloud.rs
+++ b/crates/language_models_cloud/src/language_models_cloud.rs
@@ -19,13 +19,13 @@ use http_client::{
     AsyncBody, HttpClient, HttpClientWithUrl, HttpRequestExt, Method, Response, StatusCode,
 };
 use language_model::{
-    ANTHROPIC_PROVIDER_ID, ANTHROPIC_PROVIDER_NAME, BASETEN_PROVIDER_ID, BASETEN_PROVIDER_NAME,
-    CompactionResult, DisabledReason, GOOGLE_PROVIDER_ID, GOOGLE_PROVIDER_NAME, LanguageModel,
-    LanguageModelCompletionError, LanguageModelCompletionEvent, LanguageModelEffortLevel,
-    LanguageModelId, LanguageModelName, LanguageModelProviderId, LanguageModelProviderName,
-    LanguageModelRequest, LanguageModelToolChoice, OPEN_AI_PROVIDER_ID, OPEN_AI_PROVIDER_NAME,
-    ProviderErrorCategory, RateLimiter, X_AI_PROVIDER_ID, X_AI_PROVIDER_NAME,
-    ZED_CLOUD_PROVIDER_ID, ZED_CLOUD_PROVIDER_NAME,
+    ANTHROPIC_PROVIDER_ID, ANTHROPIC_PROVIDER_NAME, CompactionResult, DisabledReason,
+    GOOGLE_PROVIDER_ID, GOOGLE_PROVIDER_NAME, LanguageModel, LanguageModelCompletionError,
+    LanguageModelCompletionEvent, LanguageModelEffortLevel, LanguageModelId, LanguageModelName,
+    LanguageModelProviderId, LanguageModelProviderName, LanguageModelRequest,
+    LanguageModelToolChoice, OPEN_AI_PROVIDER_ID, OPEN_AI_PROVIDER_NAME, ProviderErrorCategory,
+    RateLimiter, X_AI_PROVIDER_ID, X_AI_PROVIDER_NAME, ZED_CLOUD_PROVIDER_ID,
+    ZED_CLOUD_PROVIDER_NAME,
 };
 
 use schemars::JsonSchema;
@@ -607,7 +607,6 @@ impl<TP: CloudLlmTokenProvider + 'static> LanguageModel for CloudLanguageModel<T
         use cloud_llm_client::LanguageModelProvider::*;
         match self.model.provider {
             Anthropic => ANTHROPIC_PROVIDER_ID,
-            Baseten => BASETEN_PROVIDER_ID,
             OpenAi => OPEN_AI_PROVIDER_ID,
             Google => GOOGLE_PROVIDER_ID,
             XAi => X_AI_PROVIDER_ID,
@@ -618,7 +617,6 @@ impl<TP: CloudLlmTokenProvider + 'static> LanguageModel for CloudLanguageModel<T
         use cloud_llm_client::LanguageModelProvider::*;
         match self.model.provider {
             Anthropic => ANTHROPIC_PROVIDER_NAME,
-            Baseten => BASETEN_PROVIDER_NAME,
             OpenAi => OPEN_AI_PROVIDER_NAME,
             Google => GOOGLE_PROVIDER_NAME,
             XAi => X_AI_PROVIDER_NAME,
@@ -726,8 +724,7 @@ impl<TP: CloudLlmTokenProvider + 'static> LanguageModel for CloudLanguageModel<T
             // Unreachable while the `supports_explicit_compaction` guard
             // above holds, but a provider mismatch should degrade to the
             // same unsupported error rather than panic.
-            cloud_llm_client::LanguageModelProvider::Baseten
-            | cloud_llm_client::LanguageModelProvider::Google
+            cloud_llm_client::LanguageModelProvider::Google
             | cloud_llm_client::LanguageModelProvider::XAi => async {
                 Err(LanguageModelCompletionError::Other(anyhow::anyhow!(
                     "this cloud model does not support explicit compaction"
@@ -765,7 +762,7 @@ impl<TP: CloudLlmTokenProvider + 'static> LanguageModel for CloudLanguageModel<T
 
     fn supports_split_token_display(&self) -> bool {
         use cloud_llm_client::LanguageModelProvider::*;
-        matches!(self.model.provider, Baseten | OpenAi | XAi)
+        matches!(self.model.provider, OpenAi | XAi)
     }
 
     fn telemetry_id(&self) -> String {
@@ -969,8 +966,7 @@ impl<TP: CloudLlmTokenProvider + 'static> LanguageModel for CloudLanguageModel<T
                 });
                 async move { Ok(future.await?.boxed()) }.boxed()
             }
-            provider @ (cloud_llm_client::LanguageModelProvider::Baseten
-            | cloud_llm_client::LanguageModelProvider::XAi) => {
+            cloud_llm_client::LanguageModelProvider::XAi => {
                 let http_client = self.http_client.clone();
                 let token_provider = self.token_provider.clone();
                 let request = match into_open_ai(
@@ -1000,7 +996,7 @@ impl<TP: CloudLlmTokenProvider + 'static> LanguageModel for CloudLanguageModel<T
                         CompletionBody {
                             thread_id,
                             prompt_id,
-                            provider,
+                            provider: cloud_llm_client::LanguageModelProvider::XAi,
                             model: request.model.clone(),
                             provider_request: serde_json::to_value(&request).map_err(|error| {
                                 LanguageModelCompletionError::SerializeRequest {
@@ -1288,7 +1284,6 @@ pub fn provider_name(
 ) -> LanguageModelProviderName {
     match provider {
         cloud_llm_client::LanguageModelProvider::Anthropic => ANTHROPIC_PROVIDER_NAME,
-        cloud_llm_client::LanguageModelProvider::Baseten => BASETEN_PROVIDER_NAME,
         cloud_llm_client::LanguageModelProvider::OpenAi => OPEN_AI_PROVIDER_NAME,
         cloud_llm_client::LanguageModelProvider::Google => GOOGLE_PROVIDER_NAME,
         cloud_llm_client::LanguageModelProvider::XAi => X_AI_PROVIDER_NAME,