Enhance the cloud model connectin and natural language

Author: Shashikant86

README.md

@@ -484,16 +484,18 @@ Connect to any LLM provider:
 # Ollama (local, free)
 /connect ollama llama3.1:8b
 
-# OpenAI
-/connect openai gpt-4
+# OpenAI (example small model)
+/connect openai gpt-5-nano
 
-# Anthropic
+# Anthropic (paid key required)
 /connect anthropic claude-3-5-sonnet-20241022
 
-# Google Gemini
-/connect gemini gemini-2.0-flash-exp
+# Google Gemini (example model)
+/connect gemini gemini-2.5-flash
 ```
 
+> 💡 **Tip:** These are just starting points. Check your provider docs for the **latest models** (for example gpt-4o / gpt‑5 family, Gemini 2.5, latest Claude Sonnet/Opus) and plug them into `/connect`.
+
 ## 🧬 GEPA Optimization
 
 DSPy Code includes real GEPA (Genetic Pareto) optimization:
@@ -640,7 +642,7 @@ DSPy Code is currently in **Beta** and under active development. While it's func
 
 <div align="center">
 
-Built with ❤️ by **[Superagentic AI](https://super-agentic.ai)**
+Brought to you by **[Superagentic AI](https://super-agentic.ai)**
 
 Special thanks to the DSPy community and all contributors!
 
@@ -668,6 +670,6 @@ It helps others discover the project and motivates us to keep improving it.
 
 <img src="https://raw.githubusercontent.com/SuperagenticAI/dspy-code/main/docs/resource/dspy-code.png" alt="DSPy Code" width="100"/>
 
-**Made with 💜 for the DSPy community**
+**Made for the DSPy community**
 
 </div>

docs/getting-started/first-program.md

@@ -120,7 +120,7 @@ class FeedbackClassifierModule(dspy.Module):
 # Configure DSPy
 import dspy
 dspy.settings.configure(
-    lm=dspy.OpenAI(model="gpt-3.5-turbo")
+    lm=dspy.OpenAI(model="gpt-5-nano")
 )
 
 # Create and use the classifier
@@ -259,6 +259,9 @@ Generate GEPA optimization code:
 /optimize
 ```
 
+!!! warning "Optimization on Cloud Models"
+    GEPA optimization can make **a large number of LLM calls**. If you're connected to a cloud provider (OpenAI, Anthropic, Gemini), be sure you understand the potential API cost and have your quotas/billing configured before running `/optimize`.
+
 This creates a complete GEPA optimization script!
 
 **What's included:**

docs/getting-started/installation.md

@@ -55,47 +55,40 @@ source .venv/bin/activate.fish
 
 ### Step 3: Install DSPy Code
 
-```bash
-# This installs into .venv/ in your project
-pip install dspy-code
-```
-
-That's it! DSPy Code is now installed in your project.
+=== "pip"
 
-### From Source
+    ```bash
+    # This installs into .venv/ in your project
+    pip install --upgrade dspy-code
+    ```
 
-If you want the latest development version:
+=== "uv"
 
-```bash
-# Clone the repository
-git clone https://github.com/superagentic-ai/dspy-code.git
-cd dspy-code
+    ```bash
+    # If you use uv, you can install dspy-code like this
+    uv pip install --upgrade dspy-code
+    ```
 
-# Install in development mode
-pip install -e .
-```
+That's it! DSPy Code is now installed in your project.
 
-## Install DSPy
+### Step 4: Install DSPy (Optional)
 
-DSPy Code requires DSPy to be installed. If you don't have it yet:
+DSPy Code will install DSPy automatically if needed, but you can install/upgrade it explicitly:
 
-```bash
-pip install dspy
-```
+=== "pip"
 
-!!! info "DSPy Version"
-    DSPy Code works with any DSPy version (2.x, 3.x, or newer). It adapts to YOUR installed version!
+    ```bash
+    pip install --upgrade dspy
+    ```
 
-### Step 4: Install DSPy (Optional)
+=== "uv"
 
-DSPy Code works with any version of DSPy you have installed:
-
-```bash
-pip install dspy
-```
+    ```bash
+    uv pip install --upgrade dspy
+    ```
 
 !!! info "DSPy Version"
-    DSPy Code adapts to YOUR installed DSPy version! It indexes your specific version for accurate code generation and Q&A.
+    DSPy Code adapts to YOUR installed DSPy version and indexes it for accurate code generation and Q&A.
 
 ## Verify Installation
 
@@ -156,27 +149,47 @@ my-dspy-project/
 
 ## Optional Dependencies
 
-DSPy Code has optional dependencies for different features. Install only what you need:
+DSPy Code has optional dependencies for different features. Install only what you need.
 
-### For OpenAI Models
+### Cloud Model Providers (via dspy-code extras)
 
-```bash
-pip install openai
-```
+Use extras so versions stay aligned with dspy-code’s tested matrix.
 
-### For Anthropic Claude
+=== "pip"
 
-```bash
-pip install anthropic
-```
+    ```bash
+    # OpenAI support
+    pip install "dspy-code[openai]"
 
-### For Google Gemini
+    # Google Gemini support
+    pip install "dspy-code[gemini]"
 
-```bash
-pip install google-generativeai
-```
+    # Anthropic (paid key required)
+    pip install "dspy-code[anthropic]"
+
+    # Or install all cloud providers at once
+    pip install "dspy-code[llm-all]"
+    ```
+
+=== "uv"
+
+    ```bash
+    # OpenAI support
+    uv pip install "dspy-code[openai]"
+
+    # Google Gemini support
+    uv pip install "dspy-code[gemini]"
+
+    # Anthropic (paid key required)
+    uv pip install "dspy-code[anthropic]"
+
+    # Or install all cloud providers at once
+    uv pip install "dspy-code[llm-all]"
+    ```
+
+> **Note:** Anthropic has discontinued free API keys. DSPy Code fully supports Claude **if you already have a paid API key**, but Anthropic integration will simply not work without one.
 
-### For Semantic Similarity Metrics
+### Semantic Similarity Metrics
 
 ```bash
 pip install sentence-transformers scikit-learn

docs/getting-started/quick-start.md

@@ -2,15 +2,47 @@
 
 Get started with DSPy Code in 5 minutes! This guide will walk you through creating your first DSPy program.
 
-## Prerequisites
+## Prerequisites (One‑Time Setup)
 
-Before starting, make sure you have:
+You only need to do this once for each new DSPy project.
 
-1. **Created your project directory**
-2. **Set up virtual environment INSIDE the project**
-3. **Installed dspy-code**
+```bash
+# 1. Create a project directory
+mkdir my-dspy-project
+cd my-dspy-project
+
+# 2. Create a virtual environment INSIDE this directory
+python -m venv .venv
+
+# 3. Activate the virtual environment
+# For bash/zsh (macOS/Linux):
+source .venv/bin/activate
+# For fish:
+source .venv/bin/activate.fish
+# On Windows:
+.venv\Scripts\activate
+
+# 4. Install dspy-code (always upgrade for latest features)
+pip install --upgrade dspy-code
 
-If you haven't done this yet, see [Installation Guide](installation.md).
+# 5. (Optional but recommended) Install provider SDKs via dspy-code extras
+
+# If you ONLY want local models (Ollama), you can skip this step.
+
+# OpenAI support
+pip install "dspy-code[openai]"
+
+# Google Gemini support
+pip install "dspy-code[gemini]"
+
+# Anthropic (paid key required)
+pip install "dspy-code[anthropic]"
+
+# Or install all cloud providers at once
+pip install "dspy-code[llm-all]"
+```
+
+For more details, see the [Installation Guide](installation.md).
 
 ## Step 1: Navigate to Your Project
 
@@ -56,7 +88,30 @@ You'll see a beautiful welcome screen with the DSPy version and helpful tips.
     💬 Or just describe what you want to build!
     ```
 
-## Step 2: Initialize Your Project
+## Step 3: Connect a Model (Required)
+
+Before you do anything else in the CLI, you **must connect to a model**. DSPy Code relies on an LLM for code generation and understanding.
+
+```bash
+# Ollama (local, free)
+/connect ollama llama3.1:8b
+
+# Or OpenAI (example small model)
+/connect openai gpt-5-nano
+
+# Or Google Gemini (example model)
+/connect gemini gemini-2.5-flash
+```
+
+> 💡 **Tip:** These are just starting points. Use the latest models your account supports (for example gpt‑4o / gpt‑5 family, Gemini 2.5, latest Claude Sonnet/Opus) for best quality.
+
+!!! warning "Cloud Cost & Optimization"
+    - When you're connected to **cloud providers** (OpenAI, Anthropic, Gemini), remember that API usage is **billed per token**.
+    - GEPA **optimization** (via `/optimize` or generated optimization scripts) can make a *lot* of LLM calls. Only run optimization if you understand the potential cost and have quotas/billing configured.
+    - For local optimization runs with larger models, we recommend at least **32 GB RAM**.
+    - For more details, see [Model Connection (Cloud & Local)](../guide/model-connection.md) and the [Optimization Guide](../guide/optimization.md).
+
+## Step 4: Initialize Your Project
 
 Inside the CLI, type:
 
@@ -73,7 +128,7 @@ This will:
 !!! tip "What's Happening?"
     The `/init` command scans your installed DSPy version and your project code. This lets DSPy Code answer questions about DSPy and understand your project!
 
-## Step 3: Generate Your First Program
+## Step 5: Generate Your First Program
 
 Now for the fun part! Just describe what you want in plain English:
 
@@ -106,7 +161,7 @@ DSPy Code will:
             return self.predictor(text=text)
     ```
 
-## Step 4: Save Your Code
+## Step 6: Save Your Code
 
 Save the generated code to a file:
 
@@ -121,7 +176,7 @@ The file will be saved to your `generated/` directory.
     ✓ Code saved to: generated/sentiment.py
     ```
 
-## Step 5: Validate Your Code
+## Step 7: Validate Your Code
 
 Check if your code follows DSPy best practices:
 
@@ -139,7 +194,7 @@ DSPy Code will check for:
 !!! tip "Quality Checks"
     The validator catches common mistakes and suggests improvements. It's like having an expert review your code!
 
-## Step 6: Run Your Program
+## Step 8: Run Your Program
 
 Test your program:
 
@@ -154,30 +209,17 @@ DSPy Code will execute your code in a safe sandbox and show you the results.
 You just:
 
 1. ✅ Started DSPy Code
-2. ✅ Initialized a project
-3. ✅ Generated a DSPy program
-4. ✅ Saved it to a file
-5. ✅ Validated the code
-6. ✅ Ran it successfully
+2. ✅ Connected a model
+3. ✅ Initialized a project
+4. ✅ Generated a DSPy program
+5. ✅ Saved it to a file
+6. ✅ Validated the code
+7. ✅ Ran it successfully
 
 ## What's Next?
 
-### Connect a Model (Optional but Recommended)
-
-For better code generation, connect to an LLM:
-
-```
-/connect ollama llama3.1:8b
-```
-
-Or use OpenAI:
-
-```
-/connect openai gpt-4
-```
-
-!!! info "Why Connect a Model?"
-    Without a model, DSPy Code uses templates. With a model, it can understand your specific requirements and generate more customized code!
+!!! info "Why Model Connection is Required"
+    DSPy Code needs an LLM to understand your requests and generate DSPy code. Without a connected model, most interactive features will not work.
 
 ### Try More Examples
 

docs/getting-started/understanding.md

@@ -345,6 +345,11 @@ GEPA (Genetic Pareto) optimizes DSPy programs by:
 4. **Selection**: Keep best versions
 5. **Iteration**: Repeat until optimal
 
+!!! warning "Optimization Cost & Resource Considerations"
+    - **Cloud models (OpenAI, Anthropic, Gemini)**: GEPA can trigger **many LLM calls**. Only run optimization if you understand the potential API cost and have billing/quotas configured appropriately.
+    - **Local runs**: For smooth optimization on local hardware, we recommend at least **32 GB RAM**, especially with larger models.
+    - Start with small budgets and fewer examples when experimenting.
+
 ### Optimization Process
 
 **1. Prepare Data:**
@@ -411,22 +416,24 @@ This feedback helps GEPA learn!
 /connect ollama llama3.1:8b
 ```
 
-**2. OpenAI:**
+**2. OpenAI (modern SDK, example small model):**
 
 ```
-/connect openai gpt-4
+/connect openai gpt-5-nano
 ```
 
-**3. Anthropic:**
+**3. Anthropic (paid only, optional):**
 
 ```
-/connect anthropic claude-3-sonnet
+/connect anthropic claude-sonnet-4.5
 ```
 
-**4. Gemini:**
+> Anthropic no longer offers free API keys. If you have a paid key, DSPy Code will work with Claude; otherwise, just skip Anthropic and use another provider.
+
+**4. Gemini (via google-genai):**
 
 ```
-/connect gemini gemini-pro
+/connect gemini gemini-2.5-flash
 ```
 
 ### Connection Process