diff --git a/crates/hermesllm/src/lib.rs b/crates/hermesllm/src/lib.rs
index 96ff21da..602336c9 100644
--- a/crates/hermesllm/src/lib.rs
+++ b/crates/hermesllm/src/lib.rs
@@ -9,7 +9,8 @@ mod tests {
 
     #[test]
     fn openai_builder() {
-        let request = OpenAIRequest::builder("gpt-3.5-turbo")
+        let request = OpenAIRequest::builder()
+            .model("gpt-3.5-turbo")
             .temperature(0.7)
             .top_p(0.9)
             .n(1)
@@ -18,16 +19,17 @@ mod tests {
             .stop(vec!["\n".to_string()])
             .presence_penalty(0.0)
             .frequency_penalty(0.0)
-            .build();
+            .build()
+            .expect("Failed to build OpenAIRequest");
 
-        assert_eq!(request.model, "gpt-3.5-turbo");
-        assert_eq!(request.temperature, Some(0.7));
-        assert_eq!(request.top_p, Some(0.9));
-        assert_eq!(request.n, Some(1));
-        assert_eq!(request.max_tokens, Some(100));
-        assert_eq!(request.stream, Some(false));
-        assert_eq!(request.stop, Some(vec!["\n".to_string()]));
-        assert_eq!(request.presence_penalty, Some(0.0));
-        assert_eq!(request.frequency_penalty, Some(0.0));
+        assert_eq!(request.base.model, "gpt-3.5-turbo");
+        assert_eq!(request.base.temperature, Some(0.7));
+        assert_eq!(request.base.top_p, Some(0.9));
+        assert_eq!(request.base.n, Some(1));
+        assert_eq!(request.base.max_tokens, Some(100));
+        assert_eq!(request.base.stream, Some(false));
+        assert_eq!(request.base.stop, Some(vec!["\n".to_string()]));
+        assert_eq!(request.base.presence_penalty, Some(0.0));
+        assert_eq!(request.base.frequency_penalty, Some(0.0));
     }
 }
diff --git a/crates/hermesllm/src/providers/groq/mod.rs b/crates/hermesllm/src/providers/groq/mod.rs
index e69de29b..84eefe46 100644
--- a/crates/hermesllm/src/providers/groq/mod.rs
+++ b/crates/hermesllm/src/providers/groq/mod.rs
@@ -0,0 +1,27 @@
+pub mod types;
+
+use thiserror::Error;
+
+use crate::providers::groq::types::{GroqRequest, GroqResponse};
+
+#[derive(Debug, Error)]
+pub enum GroqError {
+    #[error("json error: {0}")]
+    JsonParseError(#[from] serde_json::Error),
+}
+
+type Result<T> = std::result::Result<T, GroqError>;
+
+impl TryFrom<&[u8]> for GroqRequest {
+    type Error = GroqError;
+    fn try_from(bytes: &[u8]) -> Result<Self> {
+        serde_json::from_slice(bytes).map_err(GroqError::from)
+    }
+}
+
+impl TryFrom<&[u8]> for GroqResponse {
+    type Error = GroqError;
+    fn try_from(bytes: &[u8]) -> Result<Self> {
+        serde_json::from_slice(bytes).map_err(GroqError::from)
+    }
+}
diff --git a/crates/hermesllm/src/providers/mod.rs b/crates/hermesllm/src/providers/mod.rs
index c414c86b..8ffc9f57 100644
--- a/crates/hermesllm/src/providers/mod.rs
+++ b/crates/hermesllm/src/providers/mod.rs
@@ -1,9 +1,12 @@
 pub mod openai;
 pub mod groq;
+pub mod deepseek;
+pub mod common_types;
 
 /// Supported LLM providers.
 #[derive(Debug, Clone, Copy, PartialEq, Eq)]
 pub enum Provider {
-    Grok,
+    Groq,
     OpenAI,
+    DeepSeek,
 }
diff --git a/crates/hermesllm/src/providers/openai/types.rs b/crates/hermesllm/src/providers/openai/types.rs
index 6695d5a2..43893fa4 100644
--- a/crates/hermesllm/src/providers/openai/types.rs
+++ b/crates/hermesllm/src/providers/openai/types.rs
@@ -1,38 +1,15 @@
 use serde::{Deserialize, Serialize};
+use crate::providers::common_types::{ChatRequestBase, ChatResponseBase};
 
-/// Represents a request to the OpenAI API (compatible with both chat and completion endpoints).
-///
-/// Fields are based on the OpenAI API schema:
-/// https://platform.openai.com/docs/api-reference/chat/create
 #[derive(Debug, Clone, Serialize, Deserialize)]
 pub struct OpenAIRequest {
-    /// The model to use (e.g., "gpt-3.5-turbo", "gpt-4").
-    pub model: String,
-    /// The list of messages for chat endpoints (use None for completion).
-    pub messages: Option<Vec<Message>>,
-    /// Sampling temperature to use (higher values = more random).
-    pub temperature: Option<f32>,
-    /// Nucleus sampling parameter.
-    pub top_p: Option<f32>,
-    /// How many completions to generate for each prompt/message.
-    pub n: Option<u32>,
-    /// Maximum number of tokens to generate.
-    pub max_tokens: Option<u32>,
-    /// Whether to stream back partial progress.
-    pub stream: Option<bool>,
-    /// Up to 4 sequences where the API will stop generating further tokens.
-    pub stop: Option<Vec<String>>,
-    /// Penalizes new tokens based on whether they appear in the text so far.
-    pub presence_penalty: Option<f32>,
-    /// Penalizes new tokens based on their frequency in the text so far.
-    pub frequency_penalty: Option<f32>,
+    #[serde(flatten)]
+    pub base: ChatRequestBase,
 }
-
-/// Builder for `OpenAIRequest`.
 #[derive(Debug, Default, Clone)]
 pub struct OpenAIRequestBuilder {
-    model: String,
-    messages: Option<Vec<Message>>,
+    model: Option<String>,
+    messages: Option<Vec<crate::providers::common_types::Message>>,
     temperature: Option<f32>,
     top_p: Option<f32>,
     n: Option<u32>,
@@ -44,14 +21,16 @@ pub struct OpenAIRequestBuilder {
 }
 
 impl OpenAIRequestBuilder {
-    pub fn new(model: impl Into<String>) -> Self {
-        Self {
-            model: model.into(),
-            ..Default::default()
-        }
+    pub fn new() -> Self {
+        Self::default()
     }
 
-    pub fn messages(mut self, messages: Vec<Message>) -> Self {
+    pub fn model(mut self, model: impl Into<String>) -> Self {
+        self.model = Some(model.into());
+        self
+    }
+
+    pub fn messages(mut self, messages: Vec<crate::providers::common_types::Message>) -> Self {
         self.messages = Some(messages);
         self
     }
@@ -96,9 +75,10 @@ impl OpenAIRequestBuilder {
         self
     }
 
-    pub fn build(self) -> OpenAIRequest {
-        OpenAIRequest {
-            model: self.model,
+    pub fn build(self) -> Result<OpenAIRequest, &'static str> {
+        let model = self.model.ok_or("model is required")?;
+        let base = crate::providers::common_types::ChatRequestBase {
+            model,
             messages: self.messages,
             temperature: self.temperature,
             top_p: self.top_p,
@@ -108,44 +88,21 @@ impl OpenAIRequestBuilder {
             stop: self.stop,
             presence_penalty: self.presence_penalty,
             frequency_penalty: self.frequency_penalty,
-        }
+        };
+        Ok(OpenAIRequest { base })
     }
 }
 
 impl OpenAIRequest {
-    pub fn builder(model: impl Into<String>) -> OpenAIRequestBuilder {
-        OpenAIRequestBuilder::new(model)
+    pub fn builder() -> OpenAIRequestBuilder {
+        OpenAIRequestBuilder::new()
     }
 }
 
-/// Represents a message in the OpenAI chat API.
 #[derive(Debug, Clone, Serialize, Deserialize)]
-pub struct Message {
-    /// The role of the message sender ("system", "user", or "assistant").
-    pub role: String,
-    /// The content of the message.
-    pub content: String,
-}
-
-#[derive(Debug, Clone, Deserialize, Serialize)]
 pub struct OpenAIResponse {
-    pub id: String,
-    pub object: String,
-    pub created: u64,
-    pub choices: Vec<Choice>,
-    pub usage: Option<Usage>,
+    #[serde(flatten)]
+    pub base: ChatResponseBase,
 }
 
-#[derive(Debug, Clone, Deserialize, Serialize)]
-pub struct Choice {
-    pub index: u32,
-    pub message: Message,
-    pub finish_reason: Option<String>,
-}
-
-#[derive(Debug, Clone, Deserialize, Serialize)]
-pub struct Usage {
-    pub prompt_tokens: u32,
-    pub completion_tokens: u32,
-    pub total_tokens: u32,
-}
+pub use crate::providers::common_types::{Message, Choice, Usage};