diff --git a/fern/apis/prod/openapi/openapi.yaml b/fern/apis/prod/openapi/openapi.yaml
index 27aba78d1..3b8cdd66a 100644
--- a/fern/apis/prod/openapi/openapi.yaml
+++ b/fern/apis/prod/openapi/openapi.yaml
@@ -2,7 +2,7 @@ openapi: 3.0.2
 info:
   title: Superagent
   description: 🥷 Run AI-agents with an API
-  version: 0.2.29
+  version: 0.2.32
 servers:
   - url: https://api.beta.superagent.sh
 paths:
@@ -195,8 +195,8 @@ paths:
                 $ref: '#/components/schemas/HTTPValidationError'
       security:
         - HTTPBearer: []
-      x-fern-sdk-group-name: agent
       x-fern-sdk-method-name: invoke
+      x-fern-sdk-group-name: agent
   /api/v1/agents/{agent_id}/llms:
     post:
       tags:
@@ -1507,6 +1507,33 @@ paths:
                 $ref: '#/components/schemas/HTTPValidationError'
       security:
         - HTTPBearer: []
+    delete:
+      tags:
+        - Vector Database
+      summary: Delete
+      description: Delete a Vector Database
+      operationId: delete_api_v1_vector_dbs__vector_db_id__delete
+      parameters:
+        - required: true
+          schema:
+            title: Vector Db Id
+            type: string
+          name: vector_db_id
+          in: path
+      responses:
+        '200':
+          description: Successful Response
+          content:
+            application/json:
+              schema: {}
+        '422':
+          description: Validation Error
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/HTTPValidationError'
+      security:
+        - HTTPBearer: []
     patch:
       tags:
         - Vector Database
@@ -1783,6 +1810,8 @@ components:
         - TOGETHER_AI
         - ANTHROPIC
         - BEDROCK
+        - GROQ
+        - MISTRAL
       type: string
       description: An enumeration.
     OpenAiAssistantParameters:
diff --git a/libs/superagent/app/agents/base.py b/libs/superagent/app/agents/base.py
index e00786ef0..305a9cb8a 100644
--- a/libs/superagent/app/agents/base.py
+++ b/libs/superagent/app/agents/base.py
@@ -6,7 +6,7 @@
 
 from app.models.request import LLMParams as LLMParamsRequest
 from app.utils.callbacks import CustomAsyncIteratorCallbackHandler
-from prisma.enums import AgentType
+from prisma.enums import AgentType, LLMProvider
 from prisma.models import LLM, Agent
 
 
@@ -21,9 +21,21 @@ class LLMParams(BaseModel):
 class LLMData(BaseModel):
     llm: LLM
     params: LLMParams
+    model: str
 
 
 class AgentBase(ABC):
+    _input: str
+    _messages: list = []
+    prompt: Any
+    tools: Any
+    session_id: str
+    enable_streaming: bool
+    output_schema: str
+    callbacks: List[CustomAsyncIteratorCallbackHandler]
+    agent_data: Agent
+    llm_data: LLMData
+
     def __init__(
         self,
         session_id: str,
@@ -40,10 +52,6 @@ def __init__(
         self.llm_data = llm_data
         self.agent_data = agent_data
 
-    _input: str
-    prompt: Any
-    tools: Any
-
     @property
     def input(self):
         return self._input
@@ -52,6 +60,14 @@ def input(self):
     def input(self, value: str):
         self._input = value
 
+    @property
+    def messages(self):
+        return self._messages
+
+    @messages.setter
+    def messages(self, value: list):
+        self._messages = value
+
     @property
     @abstractmethod
     def prompt(self) -> Any:
@@ -95,7 +111,31 @@ def llm_data(self):
             **(params),
         }
 
-        return LLMData(llm=llm, params=LLMParams.parse_obj(options))
+        params = LLMParams(
+            temperature=options.get("temperature"),
+            max_tokens=options.get("max_tokens"),
+            aws_access_key_id=(
+                options.get("aws_access_key_id")
+                if llm.provider == LLMProvider.BEDROCK
+                else None
+            ),
+            aws_secret_access_key=(
+                options.get("aws_secret_access_key")
+                if llm.provider == LLMProvider.BEDROCK
+                else None
+            ),
+            aws_region_name=(
+                options.get("aws_region_name")
+                if llm.provider == LLMProvider.BEDROCK
+                else None
+            ),
+        )
+
+        return LLMData(
+            llm=llm,
+            params=LLMParams.parse_obj(options),
+            model=self.agent_data.llmModel or self.agent_data.metadata.get("model"),
+        )
 
     async def get_agent(self):
         if self.agent_data.type == AgentType.OPENAI_ASSISTANT:
diff --git a/libs/superagent/app/agents/langchain.py b/libs/superagent/app/agents/langchain.py
index 73754b58c..c5fc39e95 100644
--- a/libs/superagent/app/agents/langchain.py
+++ b/libs/superagent/app/agents/langchain.py
@@ -46,7 +46,7 @@ def _get_llm(self):
 
         if llm_data.llm.provider == LLMProvider.OPENAI:
             return ChatOpenAI(
-                model=LLM_MAPPING[self.agent_data.llmModel],
+                model=LLM_MAPPING[self.llm_data.model],
                 openai_api_key=llm_data.llm.apiKey,
                 streaming=self.enable_streaming,
                 callbacks=self.callbacks,
diff --git a/libs/superagent/app/agents/llm.py b/libs/superagent/app/agents/llm.py
index 647a92760..9836e1fec 100644
--- a/libs/superagent/app/agents/llm.py
+++ b/libs/superagent/app/agents/llm.py
@@ -7,7 +7,7 @@
 from langchain_core.agents import AgentActionMessageLog
 from langchain_core.messages import AIMessage
 from langchain_core.utils.function_calling import convert_to_openai_function
-from litellm import acompletion, completion
+from litellm import completion, get_llm_provider, get_supported_openai_params
 
 from app.agents.base import AgentBase
 from app.tools import get_tools
@@ -25,9 +25,9 @@
 async def call_tool(
     agent_data: Agent, session_id: str, function: Any
 ) -> tuple[AgentActionMessageLog, Any]:
-    name = function.name
+    name = function.get("name")
     try:
-        args = json.loads(function.arguments)
+        args = json.loads(function.get("arguments"))
     except Exception as e:
         logger.error(f"Error parsing function arguments for {name}: {e}")
         raise e
@@ -44,33 +44,46 @@ async def call_tool(
     if not tool_to_call:
         raise Exception(f"Function {name} not found in tools")
 
-    res = await tool_to_call._arun(**args)
-
-    return (
-        AgentActionMessageLog(
-            tool=name,
-            tool_input=args,
-            log=f"\nInvoking: `{name}` with `{args}`\n\n\n",
-            message_log=[
-                AIMessage(
-                    content="",
-                    additional_kwargs={
-                        "function_call": {
-                            "arguments": args,
-                            "name": name,
-                        }
-                    },
-                )
-            ],
-        ),
-        res,
+    logging.info(f"Calling tool {name} with arguments {args}")
+
+    action_log = AgentActionMessageLog(
+        tool=name,
+        tool_input=args,
+        log=f"\nInvoking: `{name}` with `{args}`\n\n\n",
+        message_log=[
+            AIMessage(
+                content="",
+                additional_kwargs={
+                    "function_call": {
+                        "arguments": args,
+                        "name": name,
+                    }
+                },
+            )
+        ],
     )
+    tool_res = None
+    try:
+        tool_res = await tool_to_call._arun(**args)
+        logging.info(f"Tool {name} returned {tool_res}")
+    except Exception as e:
+        tool_res = f"Error calling {tool_to_call.name} tool with arguments {args}: {e}"
+        logging.error(f"Error calling tool {name}: {e}")
+
+    return (action_log, tool_res, tool_to_call.return_direct)
 
 
 class LLMAgent(AgentBase):
     @property
     def tools(self):
-        pass
+        tools = get_tools(
+            agent_data=self.agent_data,
+            session_id=self.session_id,
+        )
+        return [
+            {"type": "function", "function": convert_to_openai_function(tool)}
+            for tool in tools
+        ]
 
     @property
     def prompt(self):
@@ -91,8 +104,163 @@ def prompt(self):
         return prompt
 
     @property
-    def messages(self):
-        return [
+    def _is_tool_calling_supported(self):
+        (model, custom_llm_provider, _, _) = get_llm_provider(self.llm_data.model)
+        supported_params = get_supported_openai_params(
+            model=model, custom_llm_provider=custom_llm_provider
+        )
+
+        return "tools" in supported_params
+
+    async def _stream_by_lines(self, output: str):
+        output_by_lines = output.split("\n")
+        if len(output_by_lines) > 1:
+            for line in output_by_lines:
+                await self.streaming_callback.on_llm_new_token(line)
+                await self.streaming_callback.on_llm_new_token("\n")
+        else:
+            await self.streaming_callback.on_llm_new_token(output_by_lines[0])
+
+    async def get_agent(self):
+        if self._is_tool_calling_supported:
+            logger.info("Using native function calling")
+            return AgentExecutor(**self.__dict__)
+
+        return AgentExecutorOpenAIFunc(**self.__dict__)
+
+
+class AgentExecutor(LLMAgent):
+    """Agent Executor for LLM (with native function calling)"""
+
+    NOT_TOOLS_STREAMING_SUPPORTED_PROVIDERS = [
+        LLMProvider.GROQ,
+        LLMProvider.BEDROCK,
+    ]
+
+    intermediate_steps = []
+
+    async def _execute_tool_calls(self, tool_calls: list[dict], **kwargs):
+        messages: list = kwargs.get("messages")
+        for tool_call in tool_calls:
+            intermediate_step = await call_tool(
+                agent_data=self.agent_data,
+                session_id=self.session_id,
+                function=tool_call.get("function"),
+            )
+            (action_log, tool_res, return_direct) = intermediate_step
+            self.intermediate_steps.append((action_log, tool_res))
+            new_message = {
+                "role": "tool",
+                "name": tool_call.get("function").get("name"),
+                "content": tool_res,
+            }
+            if tool_call.get("id"):
+                new_message["tool_call_id"] = tool_call.get("id")
+
+            messages.append(new_message)
+            if return_direct:
+                if self.enable_streaming:
+                    await self._stream_by_lines(tool_res)
+                    self.streaming_callback.done.set()
+                return tool_res
+
+        self.messages = messages
+        kwargs["messages"] = self.messages
+        return await self._completion(**kwargs)
+
+    def _cleanup_output(self, output):
+        # anthropic returns a XML formatted response
+        # we need to get the content between <result> </result> tags
+        if self.llm_data.llm.provider == LLMProvider.ANTHROPIC:
+            from xmltodict import parse as xml_parse
+
+            xml_output = "<root>" + output + "</root>"
+            output = xml_parse(xml_output)
+            output = output["root"]
+
+            if isinstance(output, str):
+                return output
+            else:
+                if "result" in output:
+                    output = output.get("result")
+                else:
+                    output = output.get("#text")
+        return output
+
+    def _transform_completion_to_streaming(self, res, **kwargs):
+        # hacky way to convert non-streaming response to streaming response
+        if not kwargs.get("stream"):
+            for choice in res.choices:
+                choice.delta = choice.message
+            res = [res]
+        return res
+
+    async def _completion(self, **kwargs) -> Any:
+        logger.info(f"Calling LLM with kwargs: {kwargs}")
+        new_messages = self.messages
+
+        if kwargs.get("stream"):
+            await self.streaming_callback.on_llm_start()
+
+        should_stream_directly = (
+            self.enable_streaming
+            and self.llm_data.llm.provider
+            not in self.NOT_TOOLS_STREAMING_SUPPORTED_PROVIDERS
+            and self.llm_data.llm.provider != LLMProvider.ANTHROPIC
+        )
+
+        # TODO: Remove this when Groq and Bedrock supports streaming with tools
+        if self.llm_data.llm.provider in self.NOT_TOOLS_STREAMING_SUPPORTED_PROVIDERS:
+            logger.info(
+                f"Disabling streaming for {self.llm_data.llm.provider}, as tools are used"
+            )
+            kwargs["stream"] = False
+
+        res = completion(**kwargs)
+        res = self._transform_completion_to_streaming(res, **kwargs)
+
+        tool_calls = []
+        output = ""
+
+        for chunk in res:
+            new_message = chunk.choices[0].delta.dict()
+            # clean up tool calls
+            if new_message.get("tool_calls"):
+                new_message["role"] = "assistant"
+                new_tool_calls = new_message.get("tool_calls", [])
+                for tool_call in new_tool_calls:
+                    tool_call["type"] = "function"
+                    if "index" in tool_call:
+                        del tool_call["index"]
+
+                new_messages.append(new_message)
+                tool_calls.extend(new_tool_calls)
+
+            content = new_message.get("content", "")
+
+            if content:
+                output += content
+                if should_stream_directly:
+                    await self.streaming_callback.on_llm_new_token(content)
+
+        self.messages = new_messages
+
+        if tool_calls:
+            return await self._execute_tool_calls(tool_calls, **kwargs)
+
+        output = self._cleanup_output(output)
+
+        if not should_stream_directly:
+            await self._stream_by_lines(output)
+
+        if self.enable_streaming:
+            self.streaming_callback.done.set()
+
+        return output
+
+    async def ainvoke(self, input, *_, **kwargs):
+        self.input = input
+        self.messages = [
             {
                 "content": self.prompt,
                 "role": "system",
@@ -103,22 +271,33 @@ def messages(self):
             },
         ]
 
-    async def get_agent(self):
-        agent_executor = LLMAgentOpenAIFunctionCallingExecutor(**self.__dict__)
-        return agent_executor
+        if self.enable_streaming:
+            for callback in kwargs["config"]["callbacks"]:
+                if isinstance(callback, CustomAsyncIteratorCallbackHandler):
+                    self.streaming_callback = callback
 
+            if not self.streaming_callback:
+                raise Exception("Streaming Callback not found")
 
-class LLMAgentOpenAIFunctionCallingExecutor(LLMAgent):
-    @property
-    def tools(self):
-        tools = get_tools(
-            agent_data=self.agent_data,
-            session_id=self.session_id,
+        output = await self._completion(
+            model=self.llm_data.model,
+            api_key=self.llm_data.llm.apiKey,
+            messages=self.messages,
+            tools=self.tools if len(self.tools) > 0 else None,
+            tool_choice="auto" if len(self.tools) > 0 else None,
+            stream=self.enable_streaming,
+            **self.llm_data.params.dict(exclude_unset=True),
         )
-        return [
-            {"type": "function", "function": convert_to_openai_function(tool)}
-            for tool in tools
-        ]
+
+        return {
+            "intermediate_steps": self.intermediate_steps,
+            "input": self.input,
+            "output": output,
+        }
+
+
+class AgentExecutorOpenAIFunc(LLMAgent):
+    """Agent Executor that binded with OpenAI Function Calling"""
 
     @property
     def messages_function_calling(self):
@@ -133,10 +312,29 @@ def messages_function_calling(self):
             },
         ]
 
+    @property
+    def messages(self):
+        return [
+            {
+                "content": self.prompt,
+                "role": "system",
+            },
+            {
+                "content": self.input,
+                "role": "user",
+            },
+        ]
+
     async def ainvoke(self, input, *_, **kwargs):
         self.input = input
-        model = self.agent_data.metadata.get("model", "gpt-3.5-turbo-0125")
-        tool_responses = []
+        tool_results = []
+        if self.enable_streaming:
+            for callback in kwargs["config"]["callbacks"]:
+                if isinstance(callback, CustomAsyncIteratorCallbackHandler):
+                    self.streaming_callback = callback
+
+            if not self.streaming_callback:
+                raise Exception("Streaming Callback not found")
 
         if len(self.tools) > 0:
             openai_llm = await prisma.llm.find_first(
@@ -154,41 +352,45 @@ async def ainvoke(self, input, *_, **kwargs):
                 )
 
             res = completion(
+                api_key=openai_api_key,
                 model="gpt-3.5-turbo-0125",
                 messages=self.messages_function_calling,
                 tools=self.tools,
                 stream=False,
-                api_key=openai_api_key,
             )
 
             tool_calls = res.choices[0].message.get("tool_calls", [])
             for tool_call in tool_calls:
-                try:
-                    res = await call_tool(
-                        agent_data=self.agent_data,
-                        session_id=self.session_id,
-                        function=tool_call.function,
-                    )
-                except Exception as e:
-                    logger.error(
-                        f"Error calling function {tool_call.function.name}: {e}"
-                    )
-                    continue
-                tool_responses.append(res)
-
-        if len(tool_responses) > 0:
+                (action_log, tool_res, return_direct) = await call_tool(
+                    agent_data=self.agent_data,
+                    session_id=self.session_id,
+                    function=tool_call.function.dict(),
+                )
+                tool_results.append((action_log, tool_res))
+                if return_direct:
+                    if self.enable_streaming:
+                        await self._stream_by_lines(tool_res)
+                        self.streaming_callback.done.set()
+
+                    return {
+                        "intermediate_steps": tool_results,
+                        "input": self.input,
+                        "output": tool_res,
+                    }
+
+        if len(tool_results) > 0:
             INPUT_TEMPLATE = "{input}\n Context: {context}\n"
             self.input = INPUT_TEMPLATE.format(
                 input=self.input,
                 context="\n\n".join(
-                    [tool_response for (_, tool_response) in tool_responses]
+                    [tool_response for (_, tool_response) in tool_results]
                 ),
             )
 
         params = self.llm_data.params.dict(exclude_unset=True)
-        res = await acompletion(
+        res = completion(
             api_key=self.llm_data.llm.apiKey,
-            model=model,
+            model=self.llm_data.model,
             messages=self.messages,
             stream=self.enable_streaming,
             **params,
@@ -205,7 +407,7 @@ async def ainvoke(self, input, *_, **kwargs):
                 raise Exception("Streaming Callback not found")
             await streaming_callback.on_llm_start()
 
-            async for chunk in res:
+            for chunk in res:
                 token = chunk.choices[0].delta.content
                 if token:
                     output += token
@@ -216,7 +418,7 @@ async def ainvoke(self, input, *_, **kwargs):
             output = res.choices[0].message.content
 
         return {
-            "intermediate_steps": tool_responses,
+            "intermediate_steps": tool_results,
             "input": self.input,
             "output": output,
         }
diff --git a/libs/superagent/app/api/agents.py b/libs/superagent/app/api/agents.py
index 9204c7c74..20e7019e8 100644
--- a/libs/superagent/app/api/agents.py
+++ b/libs/superagent/app/api/agents.py
@@ -501,10 +501,14 @@ async def send_message(
                 from langchain.output_parsers.json import SimpleJsonOutputParser
 
                 parser = SimpleJsonOutputParser()
-                parsed_schema = str(parser.parse(schema_tokens))
+                try:
+                    parsed_res = parser.parse(schema_tokens)
+                except Exception as e:
+                    logger.error(f"Error parsing output: {e}")
+                    parsed_res = {}
 
                 # stream line by line to prevent streaming large data in one go
-                for line in parsed_schema.split("\n"):
+                for line in json.dumps(parsed_res).split("\n"):
                     async for val in stream_dict_keys(
                         {"event": "message", "data": line}
                     ):
@@ -603,8 +607,12 @@ async def send_message(
     if output_schema:
         from langchain.output_parsers.json import SimpleJsonOutputParser
 
-        json_parser = SimpleJsonOutputParser()
-        output["output"] = json_parser.parse(text=output["output"])
+        parser = SimpleJsonOutputParser()
+        try:
+            output["output"] = parser.parse(text=output["output"])
+        except Exception as e:
+            logger.error(f"Error parsing output: {e}")
+            output["output"] = {}
 
     return {"success": True, "data": output}
 
diff --git a/libs/superagent/app/api/workflow_configs/saml_schema.py b/libs/superagent/app/api/workflow_configs/saml_schema.py
index 837da1242..decc22b97 100644
--- a/libs/superagent/app/api/workflow_configs/saml_schema.py
+++ b/libs/superagent/app/api/workflow_configs/saml_schema.py
@@ -149,6 +149,9 @@ class LLMAgentTool(BaseAgentToolModel, LLMAgent):
     LLMProvider.TOGETHER_AI.value,
     LLMProvider.ANTHROPIC.value,
     LLMProvider.BEDROCK.value,
+    LLMProvider.GROQ.value,
+    LLMProvider.MISTRAL.value,
+    LLMProvider.COHERE_CHAT.value,
 ]
 
 
@@ -159,6 +162,9 @@ class Workflow(BaseModel):
     perplexity: Optional[LLMAgent]
     together_ai: Optional[LLMAgent]
     bedrock: Optional[LLMAgent]
+    groq: Optional[LLMAgent]
+    mistral: Optional[LLMAgent]
+    cohere_chat: Optional[LLMAgent]
     anthropic: Optional[LLMAgent]
     llm: Optional[LLMAgent] = Field(
         description="Deprecated! Use LLM providers instead. e.g. `perplexity` or `together_ai`"
diff --git a/libs/superagent/app/api/workflows.py b/libs/superagent/app/api/workflows.py
index 36c99572e..cef7ad71c 100644
--- a/libs/superagent/app/api/workflows.py
+++ b/libs/superagent/app/api/workflows.py
@@ -304,10 +304,15 @@ async def send_message() -> AsyncIterable[str]:
                         from langchain.output_parsers.json import SimpleJsonOutputParser
 
                         parser = SimpleJsonOutputParser()
-                        parsed_schema = str(parser.parse(schema_tokens))
+                        try:
+                            parsed_res = parser.parse(schema_tokens)
+                        except Exception as e:
+                            # TODO: stream schema parsing error as well
+                            logger.error(f"Error in parsing schema: {e}")
+                            parsed_res = {}
 
                         # stream line by line to prevent streaming large data in one go
-                        for line in parsed_schema.split("\n"):
+                        for line in json.dumps(parsed_res).split("\n"):
                             agent_name = workflow_step["agent_name"]
                             async for val in stream_dict_keys(
                                 {
diff --git a/libs/superagent/app/main.py b/libs/superagent/app/main.py
index bd5bfd13c..0491ecfb5 100644
--- a/libs/superagent/app/main.py
+++ b/libs/superagent/app/main.py
@@ -36,7 +36,7 @@
     title="Superagent",
     docs_url="/",
     description="🥷 Run AI-agents with an API",
-    version="0.2.29",
+    version="0.2.32",
     servers=[{"url": config("SUPERAGENT_API_URL")}],
 )
 
diff --git a/libs/superagent/app/models/tools.py b/libs/superagent/app/models/tools.py
index bc6d365f6..faa63ef62 100644
--- a/libs/superagent/app/models/tools.py
+++ b/libs/superagent/app/models/tools.py
@@ -1,6 +1,6 @@
 from typing import Optional
 
-from pydantic import BaseModel
+from pydantic import BaseModel, Field
 
 
 class AlgoliaInput(BaseModel):
@@ -61,7 +61,7 @@ class E2BCodeExecutorInput(BaseModel):
 
 
 class BrowserInput(BaseModel):
-    url: str
+    url: str = Field(..., description="A valid url including protocol to analyze")
 
 
 class GPTVisionInputModel(BaseModel):
diff --git a/libs/superagent/app/tools/__init__.py b/libs/superagent/app/tools/__init__.py
index 6c6fcebdc..5fa5a73cb 100644
--- a/libs/superagent/app/tools/__init__.py
+++ b/libs/superagent/app/tools/__init__.py
@@ -150,14 +150,14 @@ def conform_function_name(url):
     """
     Validates OpenAI function names and modifies them to conform to the regex
     """
-    regex_pattern = r"^[a-zA-Z0-9_-]{1,64}$"
+    regex_pattern = r"^[A-Za-z0-9_]{1,64}$"
 
     # Check if the URL matches the regex
     if re.match(regex_pattern, url):
         return url  # URL is already valid
     else:
         # Modify the URL to conform to the regex
-        valid_url = re.sub(r"[^a-zA-Z0-9_-]", "", url)[:64]
+        valid_url = re.sub(r"[^A-Za-z0-9_]", "", url)[:64]
         return valid_url
 
 
diff --git a/libs/superagent/app/tools/superrag.py b/libs/superagent/app/tools/superrag.py
index d312b3560..f1dfdcec1 100644
--- a/libs/superagent/app/tools/superrag.py
+++ b/libs/superagent/app/tools/superrag.py
@@ -1,3 +1,4 @@
+import json
 import logging
 
 from langchain_community.tools import BaseTool
@@ -57,7 +58,7 @@ async def _arun(
 
         credentials = get_superrag_compatible_credentials(provider.options)
 
-        return self.superrag_service.query(
+        res = self.superrag_service.query(
             {
                 "vector_database": {"type": database_provider, "config": credentials},
                 "index_name": index_name,
@@ -67,3 +68,4 @@ async def _arun(
                 "interpreter_mode": interpreter_mode,
             }
         )
+        return json.dumps(res)
diff --git a/libs/superagent/app/utils/callbacks.py b/libs/superagent/app/utils/callbacks.py
index 02171ac24..21bc63770 100644
--- a/libs/superagent/app/utils/callbacks.py
+++ b/libs/superagent/app/utils/callbacks.py
@@ -42,7 +42,7 @@ async def on_agent_finish(self, finish: AgentFinish, **_: Any) -> Any:
 
             while not self.queue.empty():
                 await asyncio.sleep(0.1)
-            self.done.set()
+        self.done.set()
 
     async def on_llm_start(self, *_: Any, **__: Any) -> None:
         # If two calls are made in a row, this resets the state
@@ -93,6 +93,7 @@ async def aiter(self) -> AsyncIterator[str]:
             if token_or_done is True:
                 continue
             self.is_stream_started = True
+
             yield token_or_done
 
 
diff --git a/libs/superagent/app/workflows/base.py b/libs/superagent/app/workflows/base.py
index 648207cf6..3ddc63eb6 100644
--- a/libs/superagent/app/workflows/base.py
+++ b/libs/superagent/app/workflows/base.py
@@ -1,3 +1,4 @@
+import logging
 from typing import Any, List
 
 from agentops.langchain_callback_handler import (
@@ -9,6 +10,8 @@
 from app.agents.base import AgentFactory
 from app.utils.callbacks import CustomAsyncIteratorCallbackHandler
 
+logger = logging.getLogger(__name__)
+
 
 class WorkflowBase:
     def __init__(
@@ -53,10 +56,14 @@ async def arun(self, input: Any):
             )
             if output_schema:
                 # TODO: throw error if output is not valid
-                json_parser = SimpleJsonOutputParser()
-                agent_response["output"] = json_parser.parse(
-                    text=agent_response["output"]
-                )
+                parser = SimpleJsonOutputParser()
+                try:
+                    agent_response["output"] = parser.parse(
+                        text=agent_response["output"]
+                    )
+                except Exception as e:
+                    logger.error(f"Error parsing output: {e}")
+                    agent_response["output"] = {}
 
             previous_output = agent_response.get("output")
             steps_output.append(agent_response)
diff --git a/libs/superagent/poetry.lock b/libs/superagent/poetry.lock
index a5f2389a3..9857cdffe 100644
--- a/libs/superagent/poetry.lock
+++ b/libs/superagent/poetry.lock
@@ -2311,13 +2311,13 @@ requests = ">=2,<3"
 
 [[package]]
 name = "litellm"
-version = "1.35.2"
+version = "1.35.21"
 description = "Library to easily interface with LLM API providers"
 optional = false
 python-versions = "!=2.7.*,!=3.0.*,!=3.1.*,!=3.2.*,!=3.3.*,!=3.4.*,!=3.5.*,!=3.6.*,!=3.7.*,>=3.8"
 files = [
-    {file = "litellm-1.35.2-py3-none-any.whl", hash = "sha256:686ee040154d7062b0078d882fa6399c5c7cc5ec9b5266490dee68f1b8905a36"},
-    {file = "litellm-1.35.2.tar.gz", hash = "sha256:062e5be75196da7348ae0c4f60d396f0b23ee874708ed81c40f7675161213385"},
+    {file = "litellm-1.35.21-py3-none-any.whl", hash = "sha256:907230b7ff57c853e32d04274c2bb01f75e77d49220bd3d4d8fa02cfe6d3492a"},
+    {file = "litellm-1.35.21.tar.gz", hash = "sha256:be0f9452fa357996e194c88eebc94f742be2fa623afd137a91b1e60ce5c3821f"},
 ]
 
 [package.dependencies]
@@ -6025,4 +6025,4 @@ testing = ["big-O", "jaraco.functools", "jaraco.itertools", "more-itertools", "p
 [metadata]
 lock-version = "2.0"
 python-versions = "^3.8.1, <3.12"
-content-hash = "c390f22730e24482e7f42cc8140d339a1025fb1e25021c6d886cd4dafab3a622"
+content-hash = "9049d2eda40cf7a7809de8eeac32efac0753e9006d50b8ef98aca4ef75f0e703"
diff --git a/libs/superagent/prisma/migrations/20240418181431_add_mistral/migration.sql b/libs/superagent/prisma/migrations/20240418181431_add_mistral/migration.sql
new file mode 100644
index 000000000..fa30cbda9
--- /dev/null
+++ b/libs/superagent/prisma/migrations/20240418181431_add_mistral/migration.sql
@@ -0,0 +1,2 @@
+-- AlterEnum
+ALTER TYPE "LLMProvider" ADD VALUE 'MISTRAL';
diff --git a/libs/superagent/prisma/migrations/20240418183001_add_groq/migration.sql b/libs/superagent/prisma/migrations/20240418183001_add_groq/migration.sql
new file mode 100644
index 000000000..62f59f9fc
--- /dev/null
+++ b/libs/superagent/prisma/migrations/20240418183001_add_groq/migration.sql
@@ -0,0 +1,2 @@
+-- AlterEnum
+ALTER TYPE "LLMProvider" ADD VALUE 'GROQ';
diff --git a/libs/superagent/prisma/migrations/20240420075553_add_cohere/migration.sql b/libs/superagent/prisma/migrations/20240420075553_add_cohere/migration.sql
new file mode 100644
index 000000000..3cb444fb4
--- /dev/null
+++ b/libs/superagent/prisma/migrations/20240420075553_add_cohere/migration.sql
@@ -0,0 +1,2 @@
+-- AlterEnum
+ALTER TYPE "LLMProvider" ADD VALUE 'COHERE_CHAT';
\ No newline at end of file
diff --git a/libs/superagent/prisma/schema.prisma b/libs/superagent/prisma/schema.prisma
index 6f03bca21..c3125752d 100644
--- a/libs/superagent/prisma/schema.prisma
+++ b/libs/superagent/prisma/schema.prisma
@@ -24,6 +24,9 @@ enum LLMProvider {
   TOGETHER_AI
   ANTHROPIC
   BEDROCK
+  GROQ
+  MISTRAL
+  COHERE_CHAT
 }
 
 enum LLMModel {
diff --git a/libs/superagent/pyproject.toml b/libs/superagent/pyproject.toml
index 63c75df77..7854e3041 100644
--- a/libs/superagent/pyproject.toml
+++ b/libs/superagent/pyproject.toml
@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "superagent"
-version = "0.2.29"
+version = "0.2.32"
 description = "🥷 Run AI-agents with an API"
 authors = ["Ismail Pelaseyed"]
 readme = "../../README.md"
@@ -50,7 +50,7 @@ openai = "^1.1.1"
 langchain-experimental = "^0.0.37"
 pydub = "^0.25.1"
 algoliasearch = "^3.0.0"
-litellm = "1.35.2"
+litellm = "1.35.21"
 weaviate-client = "^3.25.3"
 qdrant-client = "^1.6.9"
 vecs = "^0.4.2"
diff --git a/libs/ui/app/integrations/client-page.tsx b/libs/ui/app/integrations/client-page.tsx
index bd6e3c976..e37568a7f 100644
--- a/libs/ui/app/integrations/client-page.tsx
+++ b/libs/ui/app/integrations/client-page.tsx
@@ -15,7 +15,7 @@ export default function IntegrationsClientPage({
   configuredLLMs: any
 }) {
   return (
-    <Tabs defaultValue="storage" className="flex-1 space-y-0 overflow-hidden">
+    <Tabs defaultValue="storage" className="flex-1 space-y-0">
       <TabsList className="px-6 py-1.5">
         <TabsTrigger value="storage" className="space-x-1">
           <span>STORAGE</span>
diff --git a/libs/ui/app/integrations/llm.tsx b/libs/ui/app/integrations/llm.tsx
index 8f2bf5ba3..c375763cb 100644
--- a/libs/ui/app/integrations/llm.tsx
+++ b/libs/ui/app/integrations/llm.tsx
@@ -54,6 +54,25 @@ const antrophicSchema = z.object({
   apiKey: z.string().nonempty("API key is required"),
   options: z.object({}),
 })
+
+const groqSchema = z.object({
+  llmType: z.literal(LLMProvider.GROQ),
+  apiKey: z.string().nonempty("API key is required"),
+  options: z.object({}),
+})
+
+const mistralSchema = z.object({
+  llmType: z.literal(LLMProvider.MISTRAL),
+  apiKey: z.string().nonempty("API key is required"),
+  options: z.object({}),
+})
+
+const cohereSchema = z.object({
+  llmType: z.literal(LLMProvider.COHERE_CHAT),
+  apiKey: z.string().nonempty("API key is required"),
+  options: z.object({}),
+})
+
 const amazonBedrockSchema = z.object({
   llmType: z.literal(LLMProvider.BEDROCK),
   apiKey: z.literal(""),
@@ -79,6 +98,9 @@ const formSchema = z.discriminatedUnion("llmType", [
   perplexityAiSchema,
   togetherAiSchema,
   antrophicSchema,
+  groqSchema,
+  mistralSchema,
+  cohereSchema,
   amazonBedrockSchema,
   azureOpenAiSchema,
 ])
diff --git a/libs/ui/config/saml.ts b/libs/ui/config/saml.ts
index 9e92d2045..156c3298c 100644
--- a/libs/ui/config/saml.ts
+++ b/libs/ui/config/saml.ts
@@ -7,7 +7,7 @@ workflows:
       name: Browser assistant 
       intro: |- 
         👋 Hi there! How can I help search for answers on the internet.
-      prompt: Use the browser to answer any questions
+      prompt: Use the browser tool to answer any questions
       tools:
         - browser:
             name: browser
@@ -22,7 +22,7 @@ workflows:
   - superagent: 
       name: Browser assistant
       llm: gpt-3.5-turbo-16k-0613
-      prompt: Use the browser to answer all questions
+      prompt: Use the browser tool to answer all questions
       intro: 👋 Hi there! How can I help you?
       tools:
         - browser:
diff --git a/libs/ui/config/site.ts b/libs/ui/config/site.ts
index 3d2088966..f65099579 100644
--- a/libs/ui/config/site.ts
+++ b/libs/ui/config/site.ts
@@ -500,6 +500,19 @@ export const siteConfig = {
         },
       ],
     },
+    {
+      disabled: false,
+      formDescription: "Please enter your Groq API key.",
+      provider: LLMProvider.GROQ,
+      name: "Groq",
+      metadata: [
+        {
+          key: "apiKey",
+          type: "input",
+          label: "Groq API Key",
+        },
+      ],
+    },
     {
       disabled: false,
       formDescription: "Please enter your AWS credentials.",
@@ -523,6 +536,32 @@ export const siteConfig = {
         },
       ],
     },
+    {
+      disabled: false,
+      formDescription: "Please enter your Mistral API key.",
+      provider: LLMProvider.MISTRAL,
+      name: "Mistral",
+      metadata: [
+        {
+          key: "apiKey",
+          type: "input",
+          label: "Mistral API Key",
+        },
+      ],
+    },
+    {
+      disabled: false,
+      formDescription: "Please enter your Cohere API key.",
+      provider: LLMProvider.COHERE_CHAT,
+      name: "Cohere",
+      metadata: [
+        {
+          key: "apiKey",
+          type: "input",
+          label: "Cohere API Key",
+        },
+      ],
+    },
     {
       disabled: false,
       formDescription: "Please enter your Azure OpenAI API key.",
diff --git a/libs/ui/models/models.ts b/libs/ui/models/models.ts
index 414c0758a..7e0cfc67f 100644
--- a/libs/ui/models/models.ts
+++ b/libs/ui/models/models.ts
@@ -4,6 +4,9 @@ export const LLMProvider = {
   TOGETHER_AI: "TOGETHER_AI",
   ANTHROPIC: "ANTHROPIC",
   BEDROCK: "BEDROCK",
+  GROQ: "GROQ",
+  MISTRAL: "MISTRAL",
+  COHERE_CHAT: "COHERE_CHAT",
   AZURE_OPENAI: "AZURE_OPENAI",
 } as const