janito 2.17.1__tar.gz → 2.19.0__tar.gz

{janito-2.17.1 → janito-2.19.0}/CHANGELOG.md

@@ -2,7 +2,33 @@
 
 All notable changes to this project will be documented in this file.
 
-## [Unreleased]
+## [2.19.0] - 2025-08-08
+
+### Added
+
+- **Alibaba Cloud Provider**: Added support for new Qwen3 1M context models
+  - `qwen3-235b-a22b-thinking-2507`: 1M context thinking model
+  - `qwen3-235b-a22b-instruct-2507`: 1M context instruct model
+  - `qwen3-30b-a3b-thinking-2507`: 1M context thinking model
+  - `qwen3-30b-a3b-instruct-2507`: 1M context instruct model
+- **Thinking Mode Support**: Added thinking mode indicators (📖) for models that support chain-of-thought reasoning
+- **Default Model Highlighting**: Added star (⭐) indicators to highlight default models in provider listings
+- **Provider Connectivity Testing**: Added `--ping` flag to test connectivity with all providers when used with `--list-providers`
+- **Enhanced Model Display**: Improved model listing with better formatting of context sizes and response limits
+
+### Changed
+
+- **Alibaba Cloud Provider**: Updated default model from `qwen3-coder-plus` to `qwen3-235b-a22b-instruct-2507` (129k context)
+- **Model Information**: Added thinking mode support tracking in LLMModelInfo class
+- **Documentation**: Updated Alibaba setup guide with comprehensive configuration instructions
+
+## [2.18.0] - 2025-08-07
+
+### Added
+
+- Added OpenAI GPT-5 models (gpt-5, gpt-5-mini, gpt-5-nano) and set default model to gpt-5
+- Chat session: improved backend hostname detection
+- Cleaned up test file
 
 ## [2.17.0] - 2025-08-05
 

{janito-2.17.1 → janito-2.19.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: janito
-Version: 2.17.1
+Version: 2.19.0
 Summary: A new Python package called janito.
 Author-email: João Pinto <janito@ikignosis.org>
 Project-URL: Homepage, https://github.com/ikignosis/janito

janito-2.19.0/docs/alibaba-setup.md

@@ -0,0 +1,70 @@
+# Alibaba Cloud Qwen Setup Guide
+
+This guide explains how to set up the Alibaba Cloud Qwen provider for Janito.
+
+## Prerequisites
+
+- An Alibaba Cloud account
+- Access to the Alibaba Cloud Qwen model service
+- API key for authentication
+
+## Getting Your API Key
+
+1. Log in to your [Alibaba Cloud account](https://account.alibabacloud.com)
+2. Navigate to the **Qwen Model Service** in the console
+3. Go to **API Keys** or **Security Settings**
+4. Create a new API key or use an existing one
+5. Copy the API key value
+
+## Setting Up the API Key
+
+You can set your API key using the Janito CLI:
+
+```bash
+janito --set-api-key alibaba:your-api-key-here
+```
+
+Alternatively, you can set the environment variable:
+
+```bash
+export ALIBABA_API_KEY=your-api-key-here
+```
+
+## Configuration
+
+The Alibaba provider uses the following environment variable:
+
+- `ALIBABA_API_KEY`: Your Alibaba Cloud API key
+
+## Available Models
+
+The Alibaba provider supports the following models:
+
+- `qwen-turbo`: Fast, lightweight model for simple tasks
+- `qwen-plus`: Balanced performance and capability
+- `qwen-max`: Most capable model for complex tasks
+- `qwen3-coder-plus`: Coding-focused model with 128k context
+- `qwen3-coder-480b-a35b-instruct`: Advanced coding model
+- `qwen3-235b-a22b-thinking-2507`: 1M context thinking model
+- `qwen3-235b-a22b-instruct-2507`: 1M context instruct model
+- `qwen3-30b-a3b-thinking-2507`: 1M context thinking model
+- `qwen3-30b-a3b-instruct-2507`: 1M context instruct model
+
+## Default Model
+
+The default model is `qwen3-235b-a22b-instruct-2507`, which provides 129k context and is suitable for general-purpose tasks.
+
+## Troubleshooting
+
+**Q: I'm getting authentication errors**
+A: Verify your API key is correct and has the necessary permissions in the Alibaba Cloud console.
+
+**Q: The model is not responding**
+A: Check your internet connection and verify the Alibaba Cloud Qwen service is available in your region.
+
+**Q: I want to use a different model**
+A: Use the `--model` flag when running Janito:
+
+```bash
+janito --provider alibaba --model qwen-max
+```

{janito-2.17.1 → janito-2.19.0}/docs/cerebras-setup.md

@@ -42,6 +42,23 @@ janito -p cerebras -m qwen-3-coder-480b "Explain quantum computing"
 
 Cerebras offers several models through their API:
 
+**Production Models:**
+
+- `llama-4-scout-17b-16e-instruct`
+- `llama-3.3-70b`
+- `llama3.1-8b`
+- `qwen-3-32b`
+
+**Preview Models:**
+
+- `llama-4-maverick-17b-128e-instruct`
+- `qwen-3-235b-a22b-instruct-2507`
+- `qwen-3-235b-a22b-thinking-2507`
+- `qwen-3-coder-480b`
+- `gpt-oss-120b`
+
+**Notes:**
+
 - `qwen-3-coder-480b`: 32k context, reasoning-focused model with function calling support
 
 ## Troubleshooting

janito-2.19.0/docs/openai-setup.md

@@ -0,0 +1,85 @@
+# OpenAI Setup Guide
+
+This guide will help you set up Janito to work with OpenAI's models.
+
+## Prerequisites
+
+1. An OpenAI account
+2. An API key from OpenAI
+
+## Getting an API Key
+
+1. Go to [OpenAI's API Keys page](https://platform.openai.com/api-keys)
+2. Sign in to your OpenAI account
+3. Click on "Create new secret key"
+4. Copy the generated key and save it in a secure location
+
+## Configuration
+
+You can configure your OpenAI API key in several ways:
+
+### Option 1: Using the CLI
+
+```bash
+janito --set-api-key openai YOUR_API_KEY
+```
+
+### Option 2: Environment Variable
+
+Set the `OPENAI_API_KEY` environment variable:
+
+```bash
+export OPENAI_API_KEY=YOUR_API_KEY
+```
+
+On Windows:
+```cmd
+set OPENAI_API_KEY=YOUR_API_KEY
+```
+
+### Option 3: Configuration File
+
+Add the following to your Janito configuration file:
+
+```yaml
+providers:
+  openai:
+    api_key: YOUR_API_KEY
+```
+
+## Available Models
+
+Janito supports the following OpenAI models:
+
+- GPT-5 (default)
+- GPT-5 Mini
+- GPT-5 Nano
+- GPT-4
+- GPT-4 Turbo
+- GPT-3.5 Turbo
+
+## Usage
+
+After configuration, you can use OpenAI models with Janito:
+
+```bash
+# Use the default model (GPT-5)
+janito "Explain quantum computing"
+
+# Specify a specific model
+janito -m gpt-4 "Explain quantum computing"
+
+# Use in chat mode
+janito -c
+
+# Use in chat mode with a specific model
+janito -c -m gpt-4-turbo
+```
+
+## Troubleshooting
+
+If you encounter issues:
+
+1. Verify your API key is correct and active
+2. Check that you have sufficient credits in your OpenAI account
+3. Ensure your network connection can reach OpenAI's API endpoints

{janito-2.17.1 → janito-2.19.0}/docs/supported-providers-models.md

@@ -4,9 +4,14 @@ This page lists the supported providers and their available models.
 
 ## OpenAI
 
-- GPT-3.5 Turbo
+- GPT-5 (default)
+- GPT-5 Mini
+- GPT-5 Nano
 - GPT-4
 - GPT-4 Turbo
+- GPT-3.5 Turbo
+
+For setup instructions, see the [OpenAI Setup Guide](openai-setup.md).
 
 ## Google
 
@@ -38,6 +43,10 @@ This page lists the supported providers and their available models.
 - qwen-max
 - qwen3-coder-plus (default)
 - qwen3-coder-480b-a35b-instruct
+- qwen3-235b-a22b-thinking-2507
+- qwen3-235b-a22b-instruct-2507
+- qwen3-30b-a3b-thinking-2507
+- qwen3-30b-a3b-instruct-2507
 
 ## Z.AI
 
@@ -47,12 +56,14 @@ This page lists the supported providers and their available models.
 ## Cerebras
 
 **Production Models:**
+
 - llama-4-scout-17b-16e-instruct
 - llama-3.3-70b
 - llama3.1-8b
 - qwen-3-32b
 
 **Preview Models:**
+
 - llama-4-maverick-17b-128e-instruct
 - qwen-3-235b-a22b-instruct-2507
 - qwen-3-235b-a22b-thinking-2507
@@ -60,6 +71,7 @@ This page lists the supported providers and their available models.
 - gpt-oss-120b
 
 **Notes:**
+
 - qwen-3-coder-480b: 32k context, reasoning-focused model with function calling support
 
 <!-- Removed links to model_info.py files for all providers to avoid broken references. -->

{janito-2.17.1 → janito-2.19.0}/janito/cli/chat_mode/session.py

@@ -296,21 +296,39 @@ class ChatSession:
                 else "Unknown"
             )
 
-            # Get backend hostname if available
             backend_hostname = "Unknown"
-            if hasattr(self.agent, "driver") and self.agent.driver:
-                if hasattr(self.agent.driver, "config") and hasattr(
-                    self.agent.driver.config, "base_url"
-                ):
-                    base_url = self.agent.driver.config.base_url
-                    if base_url:
-                        try:
-                            from urllib.parse import urlparse
-
-                            parsed = urlparse(base_url)
-                            backend_hostname = parsed.netloc
-                        except Exception:
-                            backend_hostname = base_url
+            candidates = []
+            drv = getattr(self.agent, "driver", None)
+            if drv is not None:
+                cfg = getattr(drv, "config", None)
+                if cfg is not None:
+                    b = getattr(cfg, "base_url", None)
+                    if b:
+                        candidates.append(b)
+                direct_base = getattr(drv, "base_url", None)
+                if direct_base:
+                    candidates.append(direct_base)
+            cfg2 = getattr(self.agent, "config", None)
+            if cfg2 is not None:
+                b2 = getattr(cfg2, "base_url", None)
+                if b2:
+                    candidates.append(b2)
+            top_base = getattr(self.agent, "base_url", None)
+            if top_base:
+                candidates.append(top_base)
+            from urllib.parse import urlparse
+            for candidate in candidates:
+                try:
+                    if not candidate:
+                        continue
+                    parsed = urlparse(str(candidate))
+                    host = parsed.netloc or parsed.path
+                    if host:
+                        backend_hostname = host
+                        break
+                except Exception:
+                    backend_hostname = str(candidate)
+                    break
 
             self.console.print(
                 Rule(

{janito-2.17.1 → janito-2.19.0}/janito/cli/cli_commands/list_models.py

@@ -27,9 +27,23 @@ def handle_list_models(args, provider_instance):
         if models and isinstance(models[0], dict):
             _print_models_table(models, provider_name)
         else:
-            print(f"Supported models for provider '{provider_name}':")
+            # Fallback for simple string model lists
+            from rich.table import Table
+            from janito.cli.console import shared_console
+            
+            table = Table(title=f"Supported models for provider '{provider_name}'")
+            table.add_column("Model Name", style="cyan")
+            
             for m in models:
-                print(f"- {m}")
+                table.add_row(str(m))
+                
+            import sys
+            if sys.stdout.isatty():
+                shared_console.print(table)
+            else:
+                print(f"Supported models for provider '{provider_name}':")
+                for m in models:
+                    print(f"- {m}")
     except Exception as e:
         print(f"Error listing models for provider '{provider_name}': {e}")
     return

janito-2.19.0/janito/cli/cli_commands/list_providers.py

@@ -0,0 +1,15 @@
+"""
+CLI Command: List supported LLM providers
+"""
+
+from janito.provider_registry import list_providers
+from janito.cli.cli_commands.ping_providers import handle_ping_providers
+
+
+def handle_list_providers(args=None):
+    # Check if ping flag is set
+    if args and getattr(args, 'ping', False):
+        handle_ping_providers(args)
+    else:
+        list_providers()
+    return

janito-2.19.0/janito/cli/cli_commands/model_utils.py

@@ -0,0 +1,117 @@
+"""
+Utilities for model-related CLI output
+"""
+
+
+def _print_models_table(models, provider_name):
+    from rich.table import Table
+    from janito.cli.console import shared_console
+
+    table = Table(title=f"Supported models for provider '{provider_name}'")
+    table.add_column("Model Name", style="cyan")
+    table.add_column("Vendor", style="yellow", justify="center")
+    table.add_column("Context", style="magenta", justify="center")
+    table.add_column("Max Input", style="green", justify="center")
+    table.add_column("CoT", style="blue", justify="center")
+    table.add_column("Max Response", style="red", justify="center")
+    table.add_column("Thinking", style="bright_black", justify="center")
+    table.add_column("Driver", style="white")
+
+    # Get default model for this provider
+    from janito.providers.registry import LLMProviderRegistry
+    try:
+        provider_class = LLMProviderRegistry.get(provider_name)
+        default_model = getattr(provider_class, "DEFAULT_MODEL", None)
+    except:
+        default_model = None
+    
+    for m in models:
+        name = str(m.get("name", ""))
+        
+        # Highlight default model with different color
+        if name == default_model:
+            name = f"[bold green]⭐ {name}[/bold green]"
+        
+        vendor = "Open" if m.get("open") is True or m.get("open") == "Open" else "Locked"
+        
+        context = _format_context(m.get("context", ""))
+        max_input = _format_k(m.get("max_input", ""))
+        max_cot = _format_k(m.get("max_cot", ""))
+        max_response = _format_k(m.get("max_response", ""))
+        
+        # Determine thinking indicators
+        thinking_supported = m.get("thinking_supported") is True or m.get("thinking_supported") == "True"
+        cot_value = m.get("max_cot", "")
+        
+        thinking_icon = "📖" if thinking_supported and m.get("thinking", False) else ""
+        # Only show CoT value if it's a valid number and thinking is supported
+        cot_display = ""
+        if thinking_supported and cot_value and str(cot_value).lower() != "n/a":
+            cot_display = _format_k(cot_value)
+        
+        driver = _format_driver(m.get("driver", ""))
+        
+        table.add_row(name, vendor, context, max_input, cot_display, max_response, thinking_icon, driver)
+
+    import sys
+    if sys.stdout.isatty():
+        shared_console.print(table)
+    else:
+        # ASCII-friendly fallback table when output is redirected
+        print(f"Supported models for provider '{provider_name}'")
+        print("Model Name | Vendor | Context | Max Input | CoT | Max Response | Thinking | Driver")
+        
+        # Get default model for fallback
+        from janito.providers.registry import LLMProviderRegistry
+        try:
+            provider_class = LLMProviderRegistry.get(provider_name)
+            default_model = getattr(provider_class, "DEFAULT_MODEL", None)
+        except:
+            default_model = None
+            
+        for m in models:
+            name = str(m.get("name", ""))
+            if name == default_model:
+                name = f"⭐ {name} (default)"
+            
+            vendor = "Open" if m.get("open") is True or m.get("open") == "Open" else "Locked"
+            context = _format_context(m.get("context", ""))
+            max_input = _format_k(m.get("max_input", ""))
+            max_cot = _format_k(m.get("max_cot", ""))
+            max_response = _format_k(m.get("max_response", ""))
+            thinking_supported = m.get("thinking_supported") is True or m.get("thinking_supported") == "True"
+            cot_value = m.get("max_cot", "")
+            
+            thinking = "Y" if thinking_supported and m.get("thinking", False) else ""
+            cot_display = ""
+            if thinking_supported and cot_value and str(cot_value).lower() != "n/a":
+                cot_display = _format_k(cot_value)
+            
+            driver = _format_driver(m.get("driver", ""))
+            print(f"{name} | {vendor} | {context} | {max_input} | {cot_display} | {max_response} | {thinking} | {driver}")
+
+
+def _format_k(val):
+    """Format numeric values with k suffix for thousands."""
+    try:
+        n = int(val)
+        if n >= 1000:
+            return f"{n // 1000}k"
+        return str(n)
+    except Exception:
+        return str(val)
+
+
+def _format_context(val):
+    """Format context field which might be a single value or range."""
+    if isinstance(val, (list, tuple)) and len(val) == 2:
+        return f"{_format_k(val[0])} / {_format_k(val[1])}"
+    return _format_k(val)
+
+
+def _format_driver(val):
+    """Format driver name by removing ModelDriver suffix."""
+    if isinstance(val, (list, tuple)):
+        return ", ".join(val)
+    val_str = str(val)
+    return val_str.removesuffix("ModelDriver").strip()

janito-2.19.0/janito/cli/cli_commands/ping_providers.py

@@ -0,0 +1,55 @@
+from janito.provider_registry import list_providers
+from janito.providers.registry import LLMProviderRegistry
+from janito.cli.console import shared_console
+from rich.table import Table
+from rich.progress import Progress, SpinnerColumn, TextColumn
+import asyncio
+import time
+
+
+def handle_ping_providers(args):
+    """Ping/test connectivity for all providers."""
+    try:
+        # Get all providers
+        providers = list_providers()
+        
+        # Create table for results
+        table = Table(title="Provider Connectivity Test")
+        table.add_column("Provider", style="cyan")
+        table.add_column("Status", style="magenta")
+        table.add_column("Time", style="green")
+        table.add_column("Details", style="yellow")
+        
+        # Test each provider
+        for provider_name in providers:
+            start_time = time.time()
+            try:
+                # Get provider class
+                provider_class = LLMProviderRegistry.get(provider_name)
+                provider = provider_class()
+                
+                # Test the provider (simplified - just check if we can instantiate and get models)
+                models = provider.get_models()
+                if models:
+                    status = "✓ Connected"
+                    details = f"{len(models)} models available"
+                else:
+                    status = "⚠ No models"
+                    details = "Provider reachable but no models returned"
+                
+            except Exception as e:
+                status = "✗ Failed"
+                details = str(e)
+                
+            end_time = time.time()
+            elapsed = f"{(end_time - start_time)*1000:.0f}ms"
+            
+            table.add_row(provider_name, status, elapsed, details)
+        
+        # Print results
+        shared_console.print(table)
+        
+    except Exception as e:
+        print(f"Error testing provider connectivity: {e}")
+    
+    return

{janito-2.17.1 → janito-2.19.0}/janito/cli/main_cli.py

@@ -124,6 +124,10 @@ definition = [
         ["--list-providers"],
         {"action": "store_true", "help": "List supported LLM providers"},
     ),
+    (
+        ["--ping"],
+        {"action": "store_true", "help": "Ping/test connectivity for all providers (use with --list-providers)"},
+    ),
     (
         ["--list-drivers"],
         {
@@ -238,6 +242,19 @@ GETTER_KEYS = [
     "list_drivers",
     "region_info",
     "list_providers_region",
+    "ping",
+]
+GETTER_KEYS = [
+    "show_config",
+    "list_providers",
+    "list_profiles",
+    "list_models",
+    "list_tools",
+    "list_config",
+    "list_drivers",
+    "region_info",
+    "list_providers_region",
+    "ping",
 ]
 
 
@@ -361,6 +378,7 @@ class JanitoCLI:
             or self.args.show_config
             or self.args.list_config
             or self.args.list_drivers
+            or self.args.ping
         ):
             self._maybe_print_verbose_provider_model()
             handle_getter(self.args)

{janito-2.17.1 → janito-2.19.0}/janito/llm/model.py

@@ -10,6 +10,7 @@ class LLMModelInfo:
     max_cot: Any = "N/A"
     max_response: Any = "N/A"
     thinking_supported: Any = "N/A"
+    thinking: bool = False
     default_temp: float = 0.2
     open: Optional[Any] = None
     category: Optional[str] = None