BeehiveInnovations · petems · May 5, 2026 · May 5, 2026 · May 5, 2026 · May 5, 2026
diff --git a/.env.example b/.env.example
@@ -4,7 +4,7 @@
 # API Keys - At least one is required
 #
 # IMPORTANT: Choose ONE approach:
-# - Native APIs (Gemini/OpenAI/XAI) for direct access
+# - Native APIs (Gemini/OpenAI/XAI/ZAI) for direct access
 # - DIAL for unified enterprise access
 # - OpenRouter for unified cloud access
 # Having multiple unified providers creates ambiguity about which serves each model.
@@ -29,6 +29,9 @@ AZURE_OPENAI_ENDPOINT=https://your-resource.openai.azure.com/
 # Get your X.AI API key from: https://console.x.ai/
 XAI_API_KEY=your_xai_api_key_here
 
+# Get your Z.AI API key from: https://z.ai/manage-apikey/apikey-list
+ZAI_API_KEY=your_zai_api_key_here
+
 # Get your DIAL API key and configure host URL
 # DIAL provides unified access to multiple AI models through a single API
 DIAL_API_KEY=your_dial_api_key_here
@@ -105,6 +108,12 @@ DEFAULT_THINKING_MODE_THINKDEEP=high
 #   - grok3           (shorthand for grok-3)
 #   - grokfast        (shorthand for grok-3-fast)
 #
+# Supported Z.AI GLM models:
+#   - glm-4.6         (200K context, reasoning/coding)
+#   - glm             (shorthand for glm-4.6)
+#   - glm-4           (shorthand for glm-4.6)
+#   - glm4.6          (shorthand for glm-4.6)
+#
 # Supported DIAL models (when available in your DIAL deployment):
 #   - o3-2025-04-16   (200K context, latest O3 release)
 #   - o4-mini-2025-04-16 (200K context, latest O4 mini)
@@ -133,6 +142,7 @@ DEFAULT_THINKING_MODE_THINKDEEP=high
 #   OPENAI_ALLOWED_MODELS=o4-mini                # Single model standardization
 #   GOOGLE_ALLOWED_MODELS=flash,pro              # Allow both Gemini models
 #   XAI_ALLOWED_MODELS=grok,grok-3-fast          # Allow both GROK variants
+#   ZAI_ALLOWED_MODELS=glm,glm-4.6               # Allow GLM aliases/models
 #   DIAL_ALLOWED_MODELS=o3,o4-mini                       # Only allow O3/O4 models via DIAL
 #   DIAL_ALLOWED_MODELS=opus-4.1,sonnet-4.1                  # Only Claude 4.1 models (without thinking)
 #   DIAL_ALLOWED_MODELS=opus-4.1-thinking,sonnet-4.1-thinking # Only Claude 4.1 with thinking mode
@@ -142,6 +152,7 @@ DEFAULT_THINKING_MODE_THINKDEEP=high
 # OPENAI_ALLOWED_MODELS=
 # GOOGLE_ALLOWED_MODELS=
 # XAI_ALLOWED_MODELS=
+# ZAI_ALLOWED_MODELS=
 # DIAL_ALLOWED_MODELS=
 
 # Optional: Custom model configuration file path

diff --git a/README.md b/README.md
@@ -15,7 +15,7 @@
 [Claude Code](https://www.anthropic.com/claude-code) · [Gemini CLI](https://github.com/google-gemini/gemini-cli) · [Codex CLI](https://github.com/openai/codex) · [Qwen Code CLI](https://qwenlm.github.io/qwen-code-docs/) · [Cursor](https://cursor.com) · _and more_
 
 **With multiple models within a single prompt:**  
-Gemini · OpenAI · Anthropic · Grok · Azure · Ollama · OpenRouter · DIAL · On-Device Model
+Gemini · OpenAI · Anthropic · Grok · GLM · Azure · Ollama · OpenRouter · DIAL · On-Device Model
 
 </div>
 
@@ -150,6 +150,7 @@ For best results when using [Codex CLI](https://developers.openai.com/codex/cli)
 - **[OpenAI](https://platform.openai.com/api-keys)** - O3, GPT-5 series
 - **[Azure OpenAI](https://learn.microsoft.com/azure/ai-services/openai/)** - Enterprise deployments of GPT-4o, GPT-4.1, GPT-5 family
 - **[X.AI](https://console.x.ai/)** - Grok models
+- **[Z.AI](https://z.ai/manage-apikey/apikey-list)** - GLM models
 - **[DIAL](https://dialx.ai/)** - Vendor-agnostic model access
 - **[Ollama](https://ollama.ai/)** - Local models (free)
 
@@ -381,7 +382,7 @@ DISABLED_TOOLS=
 - **[Context revival](docs/context-revival.md)** - Continue conversations even after context resets
 
 **Model Support**
-- **Multiple providers** - Gemini, OpenAI, Azure, X.AI, OpenRouter, DIAL, Ollama
+- **Multiple providers** - Gemini, OpenAI, Azure, X.AI, Z.AI, OpenRouter, DIAL, Ollama
 - **Latest models** - GPT-5, Gemini 3.0 Pro, O3, Grok-4, local Llama
 - **[Thinking modes](docs/advanced-usage.md#thinking-modes)** - Control reasoning depth vs cost
 - **Vision support** - Analyze images, diagrams, screenshots

diff --git a/conf/zai_models.json b/conf/zai_models.json
@@ -0,0 +1,56 @@
+{
+  "_README": {
+    "description": "Model metadata for Z.AI GLM API access.",
+    "documentation": "https://github.com/BeehiveInnovations/pal-mcp-server/blob/main/docs/custom_models.md",
+    "usage": "Models listed here are exposed directly through the Z.AI provider. Aliases are case-insensitive.",
+    "field_notes": "Matches providers/shared/model_capabilities.py.",
+    "field_descriptions": {
+      "model_name": "The model identifier (e.g., 'glm-4.6')",
+      "aliases": "Array of short names users can type instead of the full model name",
+      "context_window": "Total number of tokens the model can process (input + output combined)",
+      "max_output_tokens": "Maximum number of tokens the model can generate in a single response",
+      "supports_extended_thinking": "Whether the model supports extended reasoning tokens",
+      "supports_json_mode": "Whether the model can guarantee valid JSON output",
+      "supports_function_calling": "Whether the model supports function/tool calling",
+      "supports_images": "Whether the model can process images/visual input",
+      "supports_temperature": "Whether the model accepts temperature parameter in API calls",
+      "temperature_constraint": "Temperature constraint object. Z.AI chat completions require temperature in [0.0, 1.0].",
+      "description": "Human-readable description of the model",
+      "intelligence_score": "1-20 human rating used as the primary signal for auto-mode model ordering"
+    },
+    "verification_sources": {
+      "model_overview": "https://docs.z.ai/guides/llm/glm-4.6",
+      "api_reference": "https://docs.z.ai/api-reference/llm/chat-completion",
+      "announcement": "https://z.ai/blog/glm-4.6"
+    }
+  },
+  "models": [
+    {
+      "model_name": "glm-4.6",
+      "friendly_name": "Z.AI (GLM-4.6)",
+      "aliases": [
+        "glm",
+        "glm-4",
+        "glm4.6",
+        "glm-4.6"
+      ],
+      "intelligence_score": 12,
+      "description": "GLM-4.6 (200K context) - Z.AI reasoning and coding model",
+      "context_window": 200000,
+      "max_output_tokens": 128000,
+      "supports_extended_thinking": true,
+      "supports_system_prompts": true,
+      "supports_streaming": true,
+      "supports_function_calling": true,
+      "supports_json_mode": true,
+      "supports_images": false,
+      "supports_temperature": true,
+      "temperature_constraint": {
+        "type": "range",
+        "min": 0.0,
+        "max": 1.0,
+        "default": 1.0
+      }
+    }
+  ]
+}
diff --git a/docker-compose.yml b/docker-compose.yml
@@ -25,6 +25,7 @@ services:
       - OPENAI_API_KEY=${OPENAI_API_KEY}
       - ANTHROPIC_API_KEY=${ANTHROPIC_API_KEY}
       - XAI_API_KEY=${XAI_API_KEY}
+      - ZAI_API_KEY=${ZAI_API_KEY}
       - DIAL_API_KEY=${DIAL_API_KEY}
       - DIAL_API_HOST=${DIAL_API_HOST}
       - DIAL_API_VERSION=${DIAL_API_VERSION}

diff --git a/docker/README.md b/docker/README.md
@@ -5,7 +5,7 @@
 ### 1. Prerequisites
 
 - Docker installed (Docker Compose optional)
-- At least one API key (Gemini, OpenAI, xAI, etc.)
+- At least one API key (Gemini, OpenAI, X.AI, Z.AI, etc.)
 
 ### 2. Configuration
 
@@ -14,7 +14,7 @@
 cp .env.example .env
 
 # Edit with your API keys (at least one required)
-# Required: GEMINI_API_KEY or OPENAI_API_KEY or XAI_API_KEY
+# Required: GEMINI_API_KEY or OPENAI_API_KEY or XAI_API_KEY or ZAI_API_KEY
 nano .env
 ```
 
@@ -293,6 +293,7 @@ DEFAULT_THINKING_MODE_THINKDEEP=high
 # Optional API keys (leave empty if not used)
 ANTHROPIC_API_KEY=
 XAI_API_KEY=
+ZAI_API_KEY=
 DIAL_API_KEY=
 OPENROUTER_API_KEY=
 CUSTOM_API_URL=

diff --git a/docker/scripts/deploy.ps1 b/docker/scripts/deploy.ps1
@@ -34,6 +34,7 @@ function Test-EnvironmentVariables {
         "GOOGLE_API_KEY", 
         "OPENAI_API_KEY",
         "XAI_API_KEY",
+        "ZAI_API_KEY",
         "DIAL_API_KEY",
         "OPENROUTER_API_KEY"
     )

diff --git a/docker/scripts/deploy.sh b/docker/scripts/deploy.sh
@@ -12,7 +12,7 @@ echo -e "${GREEN}=== Deploying PAL MCP Server ===${NC}"
 # Function to check if required environment variables are set
 check_env_vars() {
     # At least one of these API keys must be set
-    local required_vars=("GEMINI_API_KEY" "GOOGLE_API_KEY" "OPENAI_API_KEY" "XAI_API_KEY" "DIAL_API_KEY" "OPENROUTER_API_KEY")
+    local required_vars=("GEMINI_API_KEY" "GOOGLE_API_KEY" "OPENAI_API_KEY" "XAI_API_KEY" "ZAI_API_KEY" "DIAL_API_KEY" "OPENROUTER_API_KEY")
 
     local has_api_key=false
     for var in "${required_vars[@]}"; do

diff --git a/docker/scripts/healthcheck.py b/docker/scripts/healthcheck.py
@@ -68,6 +68,7 @@ def check_environment():
         "GOOGLE_API_KEY",
         "OPENAI_API_KEY",
         "XAI_API_KEY",
+        "ZAI_API_KEY",
         "DIAL_API_KEY",
         "OPENROUTER_API_KEY",
     ]

diff --git a/docs/advanced-usage.md b/docs/advanced-usage.md
@@ -49,6 +49,7 @@ Regardless of your default configuration, you can specify models per request:
 | **`gpt5-nano`** (GPT-5 Nano) | OpenAI | 400K tokens | Fastest, cheapest GPT-5 variant | Summarization and classification tasks |
 | **`grok-4`** | X.AI | 256K tokens | Latest flagship Grok model with reasoning, vision | Complex analysis, reasoning tasks |
 | **`grok-4.1-fast-reasoning`** | X.AI | 2M tokens | High-performance Grok 4.1 Fast Reasoning with vision | Fast responses and light reasoning |
+| **`glm-4.6`** | Z.AI | 200K tokens | GLM flagship model with extended thinking | Advanced reasoning and coding workflows |
 | **`llama`** (Llama 3.2) | Custom/Local | 128K tokens | Local inference, privacy | On-device analysis, cost-free processing |
 | **Any model** | OpenRouter | Varies | Access to GPT-4, Claude, Llama, etc. | User-specified or based on task requirements |
 
@@ -72,6 +73,7 @@ cloud models (expensive/powerful) AND local models (free/private) in the same co
   - **GPT-5 Mini**: Balanced efficiency and capability
   - **GPT-5 Nano**: Optimized for fast, low-cost tasks
 - **Grok-4 / Grok-4.1-fast-reasoning**: Extended thinking support, vision capabilities (256K / 2M context)
+- **GLM-4.6**: Z.AI flagship model with extended thinking support (200K context)
 
 ## Model Usage Restrictions
 

diff --git a/docs/configuration.md b/docs/configuration.md
@@ -37,6 +37,10 @@ OPENAI_API_KEY=your_openai_api_key_here
 # X.AI GROK API
 XAI_API_KEY=your_xai_api_key_here
 # Get from: https://console.x.ai/
+
+# Z.AI GLM API
+ZAI_API_KEY=your_zai_api_key_here
+# Get from: https://z.ai/manage-apikey/apikey-list
 ```
 
 **Option 2: OpenRouter (Access multiple models through one API)**
@@ -71,6 +75,7 @@ DEFAULT_MODEL=auto  # Claude picks best model for each task (recommended)
   - `conf/openai_models.json` – OpenAI catalogue (can be overridden with `OPENAI_MODELS_CONFIG_PATH`)
   - `conf/gemini_models.json` – Gemini catalogue (`GEMINI_MODELS_CONFIG_PATH`)
   - `conf/xai_models.json` – X.AI / GROK catalogue (`XAI_MODELS_CONFIG_PATH`)
+  - `conf/zai_models.json` – Z.AI / GLM catalogue (`ZAI_MODELS_CONFIG_PATH`)
   - `conf/openrouter_models.json` – OpenRouter catalogue (`OPENROUTER_MODELS_CONFIG_PATH`)
   - `conf/dial_models.json` – DIAL aggregation catalogue (`DIAL_MODELS_CONFIG_PATH`)
   - `conf/custom_models.json` – Custom/OpenAI-compatible endpoints (`CUSTOM_MODELS_CONFIG_PATH`)
@@ -84,6 +89,7 @@ DEFAULT_MODEL=auto  # Claude picks best model for each task (recommended)
   | OpenAI | `gpt-5.2`, `gpt-5.1-codex`, `gpt-5.1-codex-mini`, `gpt-5`, `gpt-5.2-pro`, `gpt-5-mini`, `gpt-5-nano`, `gpt-5-codex`, `gpt-4.1`, `o3`, `o3-mini`, `o3-pro`, `o4-mini` | `gpt5.2`, `gpt-5.2`, `5.2`, `gpt5.1-codex`, `codex-5.1`, `codex-mini`, `gpt5`, `gpt5pro`, `mini`, `nano`, `codex`, `o3mini`, `o3pro`, `o4mini` |
   | Gemini | `gemini-2.5-pro`, `gemini-2.5-flash`, `gemini-2.0-flash`, `gemini-2.0-flash-lite` | `pro`, `gemini-pro`, `flash`, `flash-2.0`, `flashlite` |
   | X.AI | `grok-4`, `grok-4.1-fast` | `grok`, `grok4`, `grok-4.1-fast-reasoning` |
+  | Z.AI | `glm-4.6` | `glm`, `glm-4`, `glm4.6` |
   | OpenRouter | See `conf/openrouter_models.json` for the continually evolving catalogue | e.g., `opus`, `sonnet`, `flash`, `pro`, `mistral` |
   | Custom | User-managed entries such as `llama3.2` | Define your own aliases per entry |
 
@@ -181,6 +187,9 @@ GOOGLE_ALLOWED_MODELS=flash,pro
 # X.AI GROK model restrictions
 XAI_ALLOWED_MODELS=grok-4,grok-4.1-fast-reasoning
 
+# Z.AI GLM model restrictions
+ZAI_ALLOWED_MODELS=glm-4.6
+
 # OpenRouter model restrictions (affects models via custom provider)
 OPENROUTER_ALLOWED_MODELS=opus,sonnet,mistral
 ```
@@ -209,6 +218,7 @@ GOOGLE_ALLOWED_MODELS=pro
 GOOGLE_ALLOWED_MODELS=flash,pro
 OPENAI_ALLOWED_MODELS=gpt-5.1-codex-mini,gpt-5-mini,o4-mini
 XAI_ALLOWED_MODELS=grok,grok-4.1-fast-reasoning
+ZAI_ALLOWED_MODELS=glm-4.6
 ```
 
 ### Advanced Configuration
@@ -219,6 +229,7 @@ XAI_ALLOWED_MODELS=grok,grok-4.1-fast-reasoning
 OPENAI_MODELS_CONFIG_PATH=/path/to/openai_models.json
 GEMINI_MODELS_CONFIG_PATH=/path/to/gemini_models.json
 XAI_MODELS_CONFIG_PATH=/path/to/xai_models.json
+ZAI_MODELS_CONFIG_PATH=/path/to/zai_models.json
 OPENROUTER_MODELS_CONFIG_PATH=/path/to/openrouter_models.json
 DIAL_MODELS_CONFIG_PATH=/path/to/dial_models.json
 CUSTOM_MODELS_CONFIG_PATH=/path/to/custom_models.json

diff --git a/docs/custom_models.md b/docs/custom_models.md
@@ -40,6 +40,7 @@ PAL ships multiple registries:
 - `conf/openai_models.json` – native OpenAI catalogue (override with `OPENAI_MODELS_CONFIG_PATH`)
 - `conf/gemini_models.json` – native Google Gemini catalogue (`GEMINI_MODELS_CONFIG_PATH`)
 - `conf/xai_models.json` – native X.AI / GROK catalogue (`XAI_MODELS_CONFIG_PATH`)
+- `conf/zai_models.json` – native Z.AI / GLM catalogue (`ZAI_MODELS_CONFIG_PATH`)
 - `conf/openrouter_models.json` – OpenRouter catalogue (`OPENROUTER_MODELS_CONFIG_PATH`)
 - `conf/dial_models.json` – DIAL aggregation catalogue (`DIAL_MODELS_CONFIG_PATH`)
 - `conf/custom_models.json` – local/self-hosted OpenAI-compatible catalogue (`CUSTOM_MODELS_CONFIG_PATH`)

diff --git a/docs/docker-deployment.md b/docs/docker-deployment.md
@@ -41,6 +41,9 @@ OPENAI_API_KEY=your_openai_api_key_here
 # X.AI GROK
 XAI_API_KEY=your_xai_api_key_here
 
+# Z.AI GLM
+ZAI_API_KEY=your_zai_api_key_here
+
 # OpenRouter (unified access)
 OPENROUTER_API_KEY=your_openrouter_api_key_here
 
@@ -252,7 +255,7 @@ cat .env
 **4. API key validation errors:**
 ```bash
 # Check environment variables in container
-docker exec pal-mcp-server env | grep -E "(GEMINI|OPENAI|XAI)"
+docker exec pal-mcp-server env | grep -E "(GEMINI|OPENAI|XAI|ZAI)"
 ```
 
 ### Debug Mode

diff --git a/docs/getting-started.md b/docs/getting-started.md
@@ -35,6 +35,10 @@ You need at least one API key. Choose based on your needs:
 - Visit [X.AI Console](https://console.x.ai/)
 - Generate an API key for Grok models
 
+**Z.AI (GLM):**
+- Visit [Z.AI API Keys](https://z.ai/manage-apikey/apikey-list)
+- Generate an API key for GLM-4.6 models
+
 **DIAL Platform:**
 - Visit [DIAL Platform](https://dialx.ai/)
 - Generate API key for vendor-agnostic model access
@@ -289,6 +293,7 @@ Add your API keys (at least one required):
 GEMINI_API_KEY=your-gemini-api-key-here      # For Gemini models  
 OPENAI_API_KEY=your-openai-api-key-here      # For GPT-5.2, GPT-5.1-Codex, O3
 XAI_API_KEY=your-xai-api-key-here            # For Grok models
+ZAI_API_KEY=your-zai-api-key-here            # For GLM models
 OPENROUTER_API_KEY=your-openrouter-key       # For multiple models
 
 # DIAL Platform (optional)

diff --git a/providers/__init__.py b/providers/__init__.py
@@ -8,6 +8,7 @@
 from .openrouter import OpenRouterProvider
 from .registry import ModelProviderRegistry
 from .shared import ModelCapabilities, ModelResponse
+from .zai import ZAIModelProvider
 
 __all__ = [
     "ModelProvider",
@@ -19,4 +20,5 @@
     "OpenAIModelProvider",
     "OpenAICompatibleProvider",
     "OpenRouterProvider",
+    "ZAIModelProvider",
 ]
diff --git a/providers/openai_compatible.py b/providers/openai_compatible.py
@@ -603,7 +603,7 @@ def generate_content(
         # Add any additional OpenAI-specific parameters
         # Use capabilities to filter parameters for reasoning models
         for key, value in kwargs.items():
-            if key in ["top_p", "frequency_penalty", "presence_penalty", "seed", "stop", "stream"]:
+            if key in ["top_p", "frequency_penalty", "presence_penalty", "seed", "stop", "stream", "extra_body"]:
                 # Reasoning models (those that don't support temperature) also don't support these parameters
                 if not supports_sampling and key in ["top_p", "frequency_penalty", "presence_penalty", "stream"]:
                     continue  # Skip unsupported parameters for reasoning models

diff --git a/providers/registries/__init__.py b/providers/registries/__init__.py
@@ -7,6 +7,7 @@
 from .openai import OpenAIModelRegistry
 from .openrouter import OpenRouterModelRegistry
 from .xai import XAIModelRegistry
+from .zai import ZAIModelRegistry
 
 __all__ = [
     "AzureModelRegistry",
@@ -16,4 +17,5 @@
     "OpenAIModelRegistry",
     "OpenRouterModelRegistry",
     "XAIModelRegistry",
+    "ZAIModelRegistry",
 ]
diff --git a/providers/registries/base.py b/providers/registries/base.py
@@ -12,7 +12,7 @@
 from utils.env import get_env
 from utils.file_utils import read_json_file
 
-from ..shared import ModelCapabilities, ProviderType, TemperatureConstraint
+from ..shared import ModelCapabilities, ProviderType, RangeTemperatureConstraint, TemperatureConstraint
 
 logger = logging.getLogger(__name__)
 
@@ -158,6 +158,8 @@ def _convert_entry(self, raw: dict) -> ModelCapabilities | None:
         temperature_hint = entry.get("temperature_constraint")
         if isinstance(temperature_hint, str):
             entry["temperature_constraint"] = TemperatureConstraint.create(temperature_hint)
+        elif isinstance(temperature_hint, dict):
+            entry["temperature_constraint"] = self._convert_temperature_constraint(temperature_hint)
         elif temperature_hint is None:
             entry["temperature_constraint"] = TemperatureConstraint.create("range")
 
@@ -178,6 +180,17 @@ def _convert_entry(self, raw: dict) -> ModelCapabilities | None:
     def _default_friendly_name(self, model_name: str) -> str:
         return model_name
 
+    def _convert_temperature_constraint(self, raw: dict) -> TemperatureConstraint:
+        constraint_type = raw.get("type", "range")
+        if constraint_type == "range":
+            return RangeTemperatureConstraint(
+                float(raw.get("min", 0.0)),
+                float(raw.get("max", 2.0)),
+                float(raw.get("default", 0.3)),
+            )
+
+        return TemperatureConstraint.create(constraint_type)
+
     def _extra_keys(self) -> set[str]:
         return set()
 

diff --git a/providers/registries/zai.py b/providers/registries/zai.py
@@ -0,0 +1,19 @@
+"""Registry loader for Z.AI model capabilities."""
+
+from __future__ import annotations
+
+from ..shared import ProviderType
+from .base import CapabilityModelRegistry
+
+
+class ZAIModelRegistry(CapabilityModelRegistry):
+    """Capability registry backed by ``conf/zai_models.json``."""
+
+    def __init__(self, config_path: str | None = None) -> None:
+        super().__init__(
+            env_var_name="ZAI_MODELS_CONFIG_PATH",
+            default_filename="zai_models.json",
+            provider=ProviderType.ZAI,
+            friendly_prefix="Z.AI ({model})",
+            config_path=config_path,
+        )