aixplain · ahmetgunduz · Aug 14, 2025 · Aug 14, 2025 · Aug 14, 2025
diff --git a/.github/workflows/main.yaml b/.github/workflows/main.yaml
@@ -111,6 +111,7 @@ jobs:
             echo "PIPELINES_RUN_URL=https://test-platform-api.aixplain.com/assets/pipeline/execution/run" >> $GITHUB_ENV
           fi
           echo "SLACK_TOKEN=${{ secrets.SLACK_TOKEN }}" >> $GITHUB_ENV
+          echo "HF_TOKEN=${{ secrets.HF_TOKEN }}" >> $GITHUB_ENV
 
       - name: Run Tests
         timeout-minutes: ${{ matrix.timeout }}

diff --git a/README.md b/README.md
@@ -1,71 +1,199 @@
-<img src="docs/assets/aixplain-brandmark-line.png" alt="aiXplain logo" title="aiXplain" align="right" height="132" width="85"/>
+# aiXplain SDK
 
-# aiXplain
+<div align="center">
+  <img src="docs/assets/aixplain-brandmark-line.png" alt="aiXplain logo" title="aiXplain" height="132" width="85"/>
+  <br>
+  <br>
+
+  [![Python 3.5+](https://img.shields.io/badge/python-3.5+-blue.svg)](https://www.python.org/downloads/)
+  [![License](https://img.shields.io/badge/License-Apache%202.0-blue.svg)](http://www.apache.org/licenses/LICENSE-2.0)
+  [![PyPI version](https://badge.fury.io/py/aiXplain.svg)](https://badge.fury.io/py/aiXplain)
+
+  **The professional AI SDK for developers and enterprises**
+</div>
 
-aixplain is a software development kit (SDK) for the [aiXplain](https://aixplain.com/) platform. With aixplain, developers can quickly and easily:
+## 📖 API Reference
 
-- [Discover](https://aixplain.com/platform/discovery/) aiXplain’s ever-expanding catalog of 35,000+ ready-to-use AI models and utilize them.
-- [Benchmark](https://aixplain.com/platform/benchmark/) AI systems by choosing models, datasets and metrics.
-- [Design](https://aixplain.com/platform/studio/) their own custom pipelines and run them.
-- [FineTune](https://aixplain.com/platform/finetune/) pre-trained models by tuning them using your data, enhancing their performance.
+- **Complete documentation:**
+  - [Python](https://docs.aixplain.com/api-reference/python/)
+  - [Swift](https://docs.aixplain.com/api-reference/swift/)
 
-🔎 **Find [models](https://platform.aixplain.com/discovery/models), [datasets](https://platform.aixplain.com/discovery/datasets), [metrics](https://platform.aixplain.com/discovery/metrics) on the platform.**
+## 🚀 Overview
 
-💛 Our repository is constantly evolving. With the help of the scientific community, we plan to add even more datasets, models, and metrics across domains and tasks.
+The aiXplain SDK is a comprehensive Python library that empowers developers to integrate cutting-edge AI capabilities into their applications with ease. Access thousands of AI models, build custom pipelines, and deploy intelligent solutions at scale.
 
-## Getting Started
+### ✨ Key Features
 
-### Installation
-To install the base package, simply,
+- **🔍 Discover**: Access 35,000+ ready-to-use AI models across multiple domains
+- **⚡ Benchmark**: Compare AI systems using comprehensive datasets and metrics
+- **🛠️ Design**: Create and deploy custom AI pipelines with our visual designer
+- **🎯 FineTune**: Enhance pre-trained models with your data for optimal performance
+
+## 📦 Installation
+
+### Basic Installation
 ```bash
 pip install aixplain
 ```
 
-To  install aiXplain with additional model building support:
+### With Model Building Support
 ```bash
 pip install aixplain[model-builder]
 ```
 
-###  API Key Setup
-Before you can use the aixplain SDK, you'll need to obtain an API key from our platform. For details refer this [Team API Key Guide](docs/user/api_setup.md).
+## 🔑 Authentication
 
-Once you get the API key, you'll  need to add this API key as an environment variable on your system.
+Get your API key from the aiXplain platform:
 
-#### Linux or macOS
-```bash
-export TEAM_API_KEY=YOUR_API_KEY
-```
-#### Windows
+1. Visit [platform.aixplain.com](https://platform.aixplain.com)
+2. Navigate to your API Keys section
+3. Generate a new Team API key
+
+### Set Your API Key
+
+**Linux/macOS:**
 ```bash
-set TEAM_API_KEY=YOUR_API_KEY
+export TEAM_API_KEY=your_api_key_here
 ```
-#### Jupyter Notebook
+
+**Windows:**
+```cmd
+set TEAM_API_KEY=your_api_key_here
 ```
-%env TEAM_API_KEY=YOUR_API_KEY
+
+**Jupyter Notebook:**
+```python
+%env TEAM_API_KEY=your_api_key_here
 ```
 
-### Usage
+For detailed setup instructions, visit [docs.aixplain.com/setup](https://docs.aixplain.com/setup).
 
-Let’s see how we can use aixplain to run a machine translation model. The following example shows an [English to French translation model](https://platform.aixplain.com/discovery/model/61dc52976eb5634cf06e97cc).
+## 🏃‍♂️ Quick Start
+
+### Running Your First Model
 
 ```python
 from aixplain.factories import ModelFactory
-model = ModelFactory.get("61dc52976eb5634cf06e97cc") # Get the ID of a model from our platform. 
-translation = model.run("This is a sample text") # Alternatively, you can input a public URL or provide a file path on your local machine.
+
+# Get an English to French translation model
+model = ModelFactory.get("61dc52976eb5634cf06e97cc")
+
+# Run translation
+result = model.run("Hello, how are you today?")
+print(result.data)  # "Bonjour, comment allez-vous aujourd'hui?"
+```
+
+### Building a Pipeline
+
+```python
+from aixplain.factories.pipeline_factory import PipelineFactory
+from aixplain.modules.pipeline.designer import Input
+
+pipeline = PipelineFactory.init("Multi Input-Output Pipeline")
+text_input_node = Input(data="text_input", data_types=["TEXT"], pipeline=pipeline)
+
+TRANSLATION_ASSET_ID = '60ddefbe8d38c51c5885f98a'
+translation_node = pipeline.translation(asset_id=TRANSLATION_ASSET_ID)
+
+SENTIMENT_ASSET_ID = '61728750720b09325cbcdc36'
+sentiment_node = pipeline.sentiment_analysis(asset_id=SENTIMENT_ASSET_ID)
+
+text_input_node.link(translation_node, 'input', 'text')
+translation_node.link(sentiment_node, 'data', 'text')
+
+translated_output_node = translation_node.use_output('data')
+sentiment_output_node = sentiment_node.use_output('data')
+
+pipeline.save()
+outputs = pipeline.run({
+    'text_input': 'This is example text to translate.'
+})
+
+print(outputs)
 ```
-*Check out the [explore section](docs/user/user_doc.md#explore) of our guide on Models to get the ID of your desired model*
 
-## Quick Links
+### Working with Agents
+
+```python
+from aixplain.factories import AgentFactory
+
+agent = AgentFactory.create(
+    name="Google Search Agent",
+    description="A search agent",
+    instructions="Use Google Search to answer queries.",
+    tools=[
+        AgentFactory.create_model_tool("65c51c556eb563350f6e1bb1")
+    ],
+    llm_id="669a63646eb56306647e1091"
+)
+response = agent.run("How can I help you today?")
+```
+
+## 📊 Core Modules
+
+| Module | Description | Documentation |
+|--------|-------------|---------------|
+| **Models** | Access 35,000+ AI models | [docs.aixplain.com/concepts/assets/models](https://docs.aixplain.com/concepts/assets/models/) |
+| **Pipelines** | Build custom AI workflows | [docs.aixplain.com/concepts/assets/pipelines](https://docs.aixplain.com/concepts/assets/pipelines) |
+| **Agents** | Deploy intelligent AI assistants | [docs.aixplain.com/concepts/assets/agents](https://docs.aixplain.com/concepts/assets/agents) |
+| **Datasets** | Manage and process data | [docs.aixplain.com/concepts/assets/data/overview](https://docs.aixplain.com/concepts/assets/data/overview) |
+| **Benchmarks** | Evaluate AI performance | [docs.aixplain.com/concepts/services/benchmark/benchmark-models](https://docs.aixplain.com/concepts/services/benchmark/benchmark-models) |
+| **FineTuning** | Customize models with your data | [docs.aixplain.com/concepts/services/finetune/finetune-llm](https://docs.aixplain.com/concepts/services/finetune/finetune-llm) |
+
+## 📚 Documentation
+
+Comprehensive documentation and guides are available at **[docs.aixplain.com](https://docs.aixplain.com)**:
+
+### 🎯 Getting Started
+- [**Quick Start Guide**](https://docs.aixplain.com/getting-started/) - Get up and running in minutes
+- [**API Key Setup**](https://docs.aixplain.com/getting-started/python) - Authentication and configuration
+- [**Tutorials**](https://docs.aixplain.com/tutorials/) - Build your first AI application
+
+### 📖 Core Guides
+- [**Discover**](https://docs.aixplain.com/concepts/assets/models) aiXplain’s ever-expanding catalog of 35,000+ ready-to-use AI models and utilize them.
+- [**Benchmark**](https://docs.aixplain.com/concepts/services/benchmark/benchmark-models) AI systems by choosing models, datasets and metrics.
+- [**Design**](https://docs.aixplain.com/concepts/assets/pipelines) their own custom pipelines and run them.
+- [**FineTune**](https://docs.aixplain.com/concepts/services/finetune/finetune-llm) pre-trained models by tuning them using your data, enhancing their performance.
+
+
+## 🛠️ Advanced Examples
+
+### Batch Processing
+```python
+# Process multiple inputs efficiently
+inputs = ["text1", "text2", "text3"]
+results = model.run_batch(inputs)
+```
+
+## 🤝 Community & Support
+
+- **📖 Documentation**: [docs.aixplain.com](https://docs.aixplain.com)
+- **💬 Discord Community**: [discord.gg/aixplain](https://discord.gg/aixplain)
+- **🐛 Issues**: [GitHub Issues](https://github.com/aixplain/aiXplain/issues)
+- **📧 Support**: support@aixplain.com
+- **🔄 Release Notes**: [GitHub Releases](https://github.com/aixplain/aiXplain/releases)
+
+## 🔗 Platform Links
+
+- **🏠 Platform Home**: [platform.aixplain.com](https://platform.aixplain.com)
+- **🔍 Model Discovery**: [platform.aixplain.com/discovery/models](https://platform.aixplain.com/discovery/models)
+- **📊 Datasets**: [platform.aixplain.com/discovery/datasets](https://platform.aixplain.com/discovery/datasets)
+- **📏 Metrics**: [platform.aixplain.com/discovery/metrics](https://platform.aixplain.com/discovery/metrics)
+
+## 📄 License
+
+This project is licensed under the Apache License 2.0. See the [LICENSE](LICENSE) file for details.
+
+## 🏢 About aiXplain
+
+aiXplain is the leading AI platform for building, deploying, and managing AI solutions at scale. We democratize AI by making cutting-edge models accessible to developers and enterprises worldwide.
+
+---
 
-* [Team API Key Guide](docs/user/api_setup.md)
-* [User Documentation](docs/user/user_doc.md)
-* [Developer Guide](docs/development/developer_guide.md)
-* [API Reference](https://docs.aixplain.com)
-* [Release notes](https://github.com/aixplain/aiXplain/releases)
+<div align="center">
+
+**Ready to build the future with AI?**
 
-## Support
-Raise issues for support in this repository.  
-Pull requests are welcome!
+[**Get Started →**](https://docs.aixplain.com/getting-started/) | [**Explore Models →**](https://platform.aixplain.com/discovery/models) | [**Join Community →**](https://discord.com/invite/T5dCmjRSYA)
 
-## Note
-The **aiXtend** python package was renamed to **aiXplain** from the release v0.1.1.
+</div>
diff --git a/aixplain/factories/agent_factory/__init__.py b/aixplain/factories/agent_factory/__init__.py
@@ -54,6 +54,7 @@ class AgentFactory:
     This class provides class methods for creating various types of agents and tools,
     as well as managing existing agents in the platform.
     """
+
     @classmethod
     def create(
         cls,
@@ -112,7 +113,10 @@ def create(
             "Note: In upcoming releases, `llm` will become a required parameter.",
             UserWarning,
         )
-        from aixplain.factories.agent_factory.utils import build_agent, build_tool_payload
+        from aixplain.factories.agent_factory.utils import (
+            build_agent,
+            build_tool_payload,
+        )
 
         agent = None
         url = urljoin(config.BACKEND_URL, "sdk/agents")
@@ -142,7 +146,7 @@ def create(
                 {
                     "type": "llm",
                     "description": "main",
-                    "parameters": llm.get_parameters().to_list() if llm.get_parameters() else None,
+                    "parameters": (llm.get_parameters().to_list() if llm.get_parameters() else None),
                 }
             )
             payload["llmId"] = llm.id
@@ -203,7 +207,11 @@ def create_from_dict(cls, dict: Dict) -> Agent:
 
     @classmethod
     def create_task(
-        cls, name: Text, description: Text, expected_output: Text, dependencies: Optional[List[Text]] = None
+        cls,
+        name: Text,
+        description: Text,
+        expected_output: Text,
+        dependencies: Optional[List[Text]] = None,
     ) -> AgentTask:
         """Create a new task for an agent.
 
@@ -217,7 +225,12 @@ def create_task(
         Returns:
             AgentTask: Created task object.
         """
-        return AgentTask(name=name, description=description, expected_output=expected_output, dependencies=dependencies)
+        return AgentTask(
+            name=name,
+            description=description,
+            expected_output=expected_output,
+            dependencies=dependencies,
+        )
 
     @classmethod
     def create_model_tool(
@@ -251,17 +264,28 @@ def create_model_tool(
         if supplier is not None:
             if isinstance(supplier, str):
                 for supplier_ in Supplier:
-                    if supplier.lower() in [supplier_.value["code"].lower(), supplier_.value["name"].lower()]:
+                    if supplier.lower() in [
+                        supplier_.value["code"].lower(),
+                        supplier_.value["name"].lower(),
+                    ]:
                         supplier = supplier_
                         break
             assert isinstance(supplier, Supplier), f"Supplier {supplier} is not a valid supplier"
         return ModelTool(
-            function=function, supplier=supplier, model=model, name=name, description=description, parameters=parameters
+            function=function,
+            supplier=supplier,
+            model=model,
+            name=name,
+            description=description,
+            parameters=parameters,
         )
 
     @classmethod
     def create_pipeline_tool(
-        cls, description: Text, pipeline: Union[Pipeline, Text], name: Optional[Text] = None
+        cls,
+        description: Text,
+        pipeline: Union[Pipeline, Text],
+        name: Optional[Text] = None,
     ) -> PipelineTool:
         """Create a new pipeline tool for use with an agent.
 
@@ -471,7 +495,12 @@ def list(cls) -> Dict:
             logging.info(f"Response for GET List Agents - Page Total: {page_total} / Total: {total}")
             for agent in results:
                 agents.append(build_agent(agent))
-            return {"results": agents, "page_total": page_total, "page_number": 0, "total": total}
+            return {
+                "results": agents,
+                "page_total": page_total,
+                "page_number": 0,
+                "total": total,
+            }
         else:
             error_msg = "Agent Listing Error: Please contact the administrators."
             if isinstance(resp, dict) and "message" in resp:
@@ -486,7 +515,7 @@ def get(cls, agent_id: Text, api_key: Optional[Text] = None) -> Agent:
 
         Args:
             agent_id (Text): ID of the agent to retrieve.
-            api_key (Optional[Text], optional): API key for authentication. 
+            api_key (Optional[Text], optional): API key for authentication.
                 Defaults to None, using the configured TEAM_API_KEY.
 
         Returns:

diff --git a/aixplain/factories/agent_factory/utils.py b/aixplain/factories/agent_factory/utils.py
@@ -44,9 +44,9 @@ def build_tool_payload(tool: Union[Tool, Model]):
             "id": tool.id,
             "name": tool.name,
             "description": tool.description,
-            "supplier": tool.supplier.value["code"] if isinstance(tool.supplier, Supplier) else tool.supplier,
+            "supplier": (tool.supplier.value["code"] if isinstance(tool.supplier, Supplier) else tool.supplier),
             "parameters": parameters,
-            "function": tool.function if hasattr(tool, "function") and tool.function is not None else None,
+            "function": (tool.function if hasattr(tool, "function") and tool.function is not None else None),
             "type": "model",
             "version": tool.version if hasattr(tool, "version") else None,
             "assetId": tool.id,
@@ -155,7 +155,10 @@ def build_llm(payload: Dict, api_key: Text = config.TEAM_API_KEY) -> LLM:
                     # Convert parameters list to dictionary format expected by ModelParameters
                     params_dict = {}
                     for param in tool["parameters"]:
-                        params_dict[param["name"]] = {"required": False, "value": param["value"]}
+                        params_dict[param["name"]] = {
+                            "required": False,
+                            "value": param["value"],
+                        }
                     # Create ModelParameters and set it on the LLM
                     from aixplain.modules.model.model_parameters import ModelParameters