diff --git a/qtext/config.py b/qtext/config.py
index 004bd0b..f454ecc 100644
--- a/qtext/config.py
+++ b/qtext/config.py
@@ -2,7 +2,7 @@
 
 import logging
 from pathlib import Path
-from typing import Annotated, Type
+from typing import Annotated, Literal, Type
 
 import msgspec
 
@@ -24,6 +24,7 @@ class VectorStoreConfig(msgspec.Struct, kw_only=True, frozen=True):
 
 
 class EmbeddingConfig(msgspec.Struct, kw_only=True, frozen=True):
+    client: Literal["openai", "cohere"] = "cohere"
     model_name: str = "thenlper/gte-base"
     dim: Annotated[int, msgspec.Meta(ge=1, le=65535)] = 768
     api_key: str = "fake"
diff --git a/qtext/emb_client.py b/qtext/emb_client.py
index d3f04bf..f315b92 100644
--- a/qtext/emb_client.py
+++ b/qtext/emb_client.py
@@ -1,5 +1,6 @@
 from __future__ import annotations
 
+import cohere
 import httpx
 import msgspec
 import openai
@@ -31,6 +32,18 @@ def embedding(self, text: str | list[str]) -> list[float]:
         return response.data[0].embedding
 
 
+class CohereEmbeddingClient:
+    def __init__(self, model_name: str, api_key: str):
+        self.client = cohere.Client(api_key=api_key)
+        self.model_name = model_name
+
+    @time_it
+    @embedding_histogram.time()
+    def embedding(self, text: str) -> list[float]:
+        response = self.client.embed([text], model=self.model_name)
+        return response.embeddings[0]
+
+
 class SparseEmbeddingClient:
     def __init__(self, endpoint: str, dim: int, timeout: int) -> None:
         self.dim = dim
diff --git a/qtext/engine.py b/qtext/engine.py
index bc5e137..f15bc79 100644
--- a/qtext/engine.py
+++ b/qtext/engine.py
@@ -3,7 +3,11 @@
 from time import perf_counter
 
 from qtext.config import Config
-from qtext.emb_client import EmbeddingClient, SparseEmbeddingClient
+from qtext.emb_client import (
+    CohereEmbeddingClient,
+    EmbeddingClient,
+    SparseEmbeddingClient,
+)
 from qtext.highlight_client import ENGLISH_STOPWORDS, HighlightClient
 from qtext.metrics import rerank_histogram
 from qtext.pg_client import PgVectorsClient
@@ -25,12 +29,18 @@ def __init__(self, config: Config) -> None:
         self.resp_cls = self.querier.table_type
         self.pg_client = PgVectorsClient(config.vector_store.url, querier=self.querier)
         self.highlight_client = HighlightClient(config.highlight.addr)
-        self.emb_client = EmbeddingClient(
-            model_name=config.embedding.model_name,
-            api_key=config.embedding.api_key,
-            endpoint=config.embedding.api_endpoint,
-            timeout=config.embedding.timeout,
-        )
+        if config.embedding.client == "openai":
+            self.emb_client = EmbeddingClient(
+                model_name=config.embedding.model_name,
+                api_key=config.embedding.api_key,
+                endpoint=config.embedding.api_endpoint,
+                timeout=config.embedding.timeout,
+            )
+        else:
+            self.emb_client = CohereEmbeddingClient(
+                model_name=config.embedding.model_name,
+                api_key=config.embedding.api_key,
+            )
         self.sparse_client = SparseEmbeddingClient(
             endpoint=config.sparse.addr,
             dim=config.sparse.dim,
diff --git a/tui.py b/tui.py
index 46517e6..b56508a 100644
--- a/tui.py
+++ b/tui.py
@@ -40,7 +40,7 @@ def compose(self) -> ComposeResult:
         with HorizontalScroll(id="namespace"):
             yield Label("Namespace:")
             yield Input(
-                "sparse_test",
+                "cohere_wiki",
                 placeholder="Type the namespace you want to query",
                 max_length=128,
                 id="namespace-input",