feat: Add 9 new models, remove 2 obsolete, update aliases

conf/openrouter_models.json

@@ -27,9 +27,7 @@
     {
       "model_name": "anthropic/claude-opus-4.5",
       "aliases": [
-        "opus",
-        "opus4.5",
-        "claude-opus"
+        "opus4.5"
       ],
       "context_window": 200000,
       "max_output_tokens": 64000,
@@ -41,10 +39,26 @@
       "description": "Claude Opus 4.5 - Anthropic's frontier reasoning model for complex software engineering and agentic workflows",
       "intelligence_score": 18
     },
+    {
+      "model_name": "anthropic/claude-opus-4.6",
+      "aliases": [
+        "opus",
+        "opus4.6",
+        "claude-opus"
+      ],
+      "context_window": 1000000,
+      "max_output_tokens": 128000,
+      "supports_extended_thinking": false,
+      "supports_json_mode": false,
+      "supports_function_calling": false,
+      "supports_images": true,
+      "max_image_size_mb": 5.0,
+      "description": "Claude Opus 4.6 - Anthropic's strongest model for coding, long-running professional tasks, and agentic workflows",
+      "intelligence_score": 18
+    },
     {
       "model_name": "anthropic/claude-sonnet-4.5",
       "aliases": [
-        "sonnet",
         "sonnet4.5"
       ],
       "context_window": 200000,
@@ -57,6 +71,22 @@
       "description": "Claude Sonnet 4.5 - High-performance model with exceptional reasoning and efficiency",
       "intelligence_score": 12
     },
+    {
+      "model_name": "anthropic/claude-sonnet-4.6",
+      "aliases": [
+        "sonnet",
+        "sonnet4.6"
+      ],
+      "context_window": 1000000,
+      "max_output_tokens": 128000,
+      "supports_extended_thinking": false,
+      "supports_json_mode": false,
+      "supports_function_calling": false,
+      "supports_images": true,
+      "max_image_size_mb": 5.0,
+      "description": "Claude Sonnet 4.6 - Frontier Sonnet with coding, agents, and professional task performance",
+      "intelligence_score": 13
+    },
     {
       "model_name": "anthropic/claude-opus-4.1",
       "aliases": [
@@ -104,12 +134,30 @@
     },
     {
       "model_name": "google/gemini-3-pro-preview",
+      "aliases": [
+        "gemini3.0",
+        "gemini-3.0",
+        "pro-openrouter"
+      ],
+      "context_window": 1048576,
+      "max_output_tokens": 65536,
+      "supports_extended_thinking": true,
+      "supports_json_mode": true,
+      "supports_function_calling": true,
+      "supports_images": true,
+      "max_image_size_mb": 20.0,
+      "allow_code_generation": true,
+      "description": "Google's Gemini 3.0 Pro via OpenRouter with vision",
+      "intelligence_score": 17
+    },
+    {
+      "model_name": "google/gemini-3.1-pro-preview",
       "aliases": [
         "pro",
         "gemini-pro",
         "gemini",
         "gemini3",
-        "pro-openrouter"
+        "gemini3.1"
       ],
       "context_window": 1048576,
       "max_output_tokens": 65536,
@@ -119,8 +167,8 @@
       "supports_images": true,
       "max_image_size_mb": 20.0,
       "allow_code_generation": true,
-      "description": "Google's Gemini 3.0 Pro via OpenRouter with vision",
-      "intelligence_score": 18
+      "description": "Google's Gemini 3.1 Pro - Frontier reasoning with enhanced software engineering and agentic capabilities",
+      "intelligence_score": 19
     },
     {
       "model_name": "google/gemini-2.5-pro",
@@ -171,25 +219,6 @@
       "description": "Mistral's largest model (text-only)",
       "intelligence_score": 11
     },
-    {
-      "model_name": "meta-llama/llama-3-70b",
-      "aliases": [
-        "llama",
-        "llama3",
-        "llama3-70b",
-        "llama-70b",
-        "llama3-openrouter"
-      ],
-      "context_window": 8192,
-      "max_output_tokens": 8192,
-      "supports_extended_thinking": false,
-      "supports_json_mode": false,
-      "supports_function_calling": false,
-      "supports_images": false,
-      "max_image_size_mb": 0.0,
-      "description": "Meta's Llama 3 70B model (text-only)",
-      "intelligence_score": 9
-    },
     {
       "model_name": "deepseek/deepseek-r1-0528",
       "aliases": [
@@ -208,23 +237,6 @@
       "description": "DeepSeek R1 with thinking mode - advanced reasoning capabilities (text-only)",
       "intelligence_score": 15
     },
-    {
-      "model_name": "perplexity/llama-3-sonar-large-32k-online",
-      "aliases": [
-        "perplexity",
-        "sonar",
-        "perplexity-online"
-      ],
-      "context_window": 32768,
-      "max_output_tokens": 32768,
-      "supports_extended_thinking": false,
-      "supports_json_mode": false,
-      "supports_function_calling": false,
-      "supports_images": false,
-      "max_image_size_mb": 0.0,
-      "description": "Perplexity's online model with web search (text-only)",
-      "intelligence_score": 9
-    },
     {
       "model_name": "openai/o3",
       "aliases": [
@@ -334,8 +346,7 @@
       "model_name": "openai/gpt-5.2-pro",
       "aliases": [
         "gpt5.2-pro",
-        "gpt5.2pro",
-        "gpt5pro"
+        "gpt5.2pro"
       ],
       "context_window": 400000,
       "max_output_tokens": 272000,
@@ -352,6 +363,28 @@
       "description": "GPT-5.2 Pro - Advanced reasoning model with highest quality responses (text+image input, text output only)",
       "intelligence_score": 18
     },
+    {
+      "model_name": "openai/gpt-5.4-pro",
+      "aliases": [
+        "gpt5.4-pro",
+        "gpt5.4pro",
+        "gpt5pro"
+      ],
+      "context_window": 1050000,
+      "max_output_tokens": 128000,
+      "supports_extended_thinking": true,
+      "supports_json_mode": true,
+      "supports_function_calling": true,
+      "supports_images": true,
+      "max_image_size_mb": 20.0,
+      "supports_temperature": false,
+      "temperature_constraint": "fixed",
+      "use_openai_response_api": true,
+      "default_reasoning_effort": "high",
+      "allow_code_generation": true,
+      "description": "GPT-5.4 Pro - OpenAI's most advanced model with enhanced reasoning and 1M context window",
+      "intelligence_score": 19
+    },
     {
       "model_name": "openai/gpt-5-codex",
       "aliases": [
@@ -450,6 +483,27 @@
       "description": "GPT-5.1 Codex (400K context, 128K output) - Agentic coding specialization available through the Responses API",
       "intelligence_score": 19
     },
+    {
+      "model_name": "openai/gpt-5.3-codex",
+      "aliases": [
+        "codex-5.3",
+        "gpt5.3-codex"
+      ],
+      "context_window": 400000,
+      "max_output_tokens": 128000,
+      "supports_extended_thinking": true,
+      "supports_json_mode": true,
+      "supports_function_calling": true,
+      "supports_images": true,
+      "max_image_size_mb": 20.0,
+      "supports_temperature": true,
+      "temperature_constraint": "fixed",
+      "use_openai_response_api": true,
+      "default_reasoning_effort": "high",
+      "allow_code_generation": true,
+      "description": "GPT-5.3 Codex - OpenAI's most advanced agentic coding model with frontier software engineering performance",
+      "intelligence_score": 19
+    },
     {
       "model_name": "openai/gpt-5.1-codex-mini",
       "aliases": [
@@ -507,6 +561,70 @@
       "temperature_constraint": "range",
       "description": "xAI's Grok 4.1 Fast Reasoning via OpenRouter (2M context) with vision and advanced reasoning",
       "intelligence_score": 15
+    },
+    {
+      "model_name": "deepseek/deepseek-v3.2-exp",
+      "aliases": [
+        "deepseek-v3",
+        "deepseek-v3.2",
+        "dsv3"
+      ],
+      "context_window": 163840,
+      "max_output_tokens": 65536,
+      "supports_extended_thinking": true,
+      "supports_json_mode": true,
+      "supports_function_calling": false,
+      "supports_images": false,
+      "max_image_size_mb": 0.0,
+      "description": "DeepSeek V3.2 Experimental - Strong reasoning capabilities (text-only)",
+      "intelligence_score": 16
+    },
+    {
+      "model_name": "mistralai/devstral-2512",
+      "aliases": [
+        "devstral"
+      ],
+      "context_window": 262144,
+      "max_output_tokens": 32768,
+      "supports_extended_thinking": false,
+      "supports_json_mode": true,
+      "supports_function_calling": true,
+      "supports_images": false,
+      "max_image_size_mb": 0.0,
+      "description": "Devstral 2 - Mistral's 123B parameter model specialized for agentic coding and codebase exploration",
+      "intelligence_score": 15
+    },
+    {
+      "model_name": "qwen/qwen3.5-397b-a17b",
+      "aliases": [
+        "qwen",
+        "qwen3.5"
+      ],
+      "context_window": 262144,
+      "max_output_tokens": 65536,
+      "supports_extended_thinking": true,
+      "supports_json_mode": true,
+      "supports_function_calling": true,
+      "supports_images": true,
+      "max_image_size_mb": 20.0,
+      "description": "Qwen 3.5 397B - Frontier reasoning model with vision, hybrid architecture (text+image+video input)",
+      "intelligence_score": 16
+    },
+    {
+      "model_name": "minimax/minimax-m2.5",
+      "aliases": [
+        "minimax",
+        "m2.5"
+      ],
+      "context_window": 196608,
+      "max_output_tokens": 196608,
+      "supports_extended_thinking": false,
+      "supports_json_mode": true,
+      "supports_function_calling": true,
+      "supports_images": false,
+      "max_image_size_mb": 0.0,
+      "description": "MiniMax M2.5 - SWE-Bench 80.2%, optimized for agent workflows and real-world productivity",
+      "intelligence_score": 16
     }
   ]
 }

config.py

@@ -14,9 +14,9 @@
 # These values are used in server responses and for tracking releases
 # IMPORTANT: This is the single source of truth for version and author info
 # Semantic versioning: MAJOR.MINOR.PATCH
-__version__ = "9.8.2"
+__version__ = "9.9.0"
 # Last update date in ISO format
-__updated__ = "2025-12-15"
+__updated__ = "2026-03-06"
 # Primary maintainer
 __author__ = "Fahad Gilani"
 

docs/custom_models.md

@@ -52,18 +52,22 @@ The curated defaults in `conf/openrouter_models.json` include popular entries su
 
 | Alias | Canonical Model | Highlights |
 |-------|-----------------|------------|
-| `opus`, `claude-opus` | `anthropic/claude-opus-4.1` | Flagship Claude reasoning model with vision |
-| `sonnet`, `sonnet4.5` | `anthropic/claude-sonnet-4.5` | Balanced Claude with high context window |
+| `opus`, `claude-opus` | `anthropic/claude-opus-4.6` | Latest Anthropic flagship (1M context, vision). `opus4.5` → 4.5, `opus4.1` → 4.1 |
+| `sonnet` | `anthropic/claude-sonnet-4.6` | Frontier Sonnet (1M context, vision). `sonnet4.5` → 4.5 |
 | `haiku` | `anthropic/claude-3.5-haiku` | Fast Claude option with vision |
-| `pro`, `gemini` | `google/gemini-2.5-pro` | Frontier Gemini with extended thinking |
+| `pro`, `gemini` | `google/gemini-3.1-pro-preview` | Latest Gemini Pro with 1M context, thinking. `gemini3.0` → 3.0 |
 | `flash` | `google/gemini-2.5-flash` | Ultra-fast Gemini with vision |
-| `mistral` | `mistralai/mistral-large-2411` | Frontier Mistral (text only) |
-| `llama3` | `meta-llama/llama-3-70b` | Large open-weight text model |
-| `deepseek-r1` | `deepseek/deepseek-r1-0528` | DeepSeek reasoning model |
-| `perplexity` | `perplexity/llama-3-sonar-large-32k-online` | Search-augmented model |
+| `gpt5pro` | `openai/gpt-5.4-pro` | OpenAI's most advanced reasoning (1M context). `gpt5.2-pro` → 5.2 Pro |
+| `codex-5.3`, `gpt5.3-codex` | `openai/gpt-5.3-codex` | Latest agentic coding model (Responses API) |
 | `gpt5.2`, `gpt-5.2`, `5.2` | `openai/gpt-5.2` | Flagship GPT-5.2 with reasoning and vision |
 | `gpt5.1-codex`, `codex-5.1` | `openai/gpt-5.1-codex` | Agentic coding specialization (Responses API) |
-| `codex-mini`, `gpt5.1-codex-mini` | `openai/gpt-5.1-codex-mini` | Cost-efficient Codex variant with streaming |
+| `codex-mini` | `openai/gpt-5.1-codex-mini` | Cost-efficient Codex variant with streaming |
+| `mistral` | `mistralai/mistral-large-2411` | Frontier Mistral (text only) |
+| `devstral` | `mistralai/devstral-2512` | 123B agentic coding model (262K context) |
+| `deepseek-r1` | `deepseek/deepseek-r1-0528` | DeepSeek reasoning model |
+| `deepseek-v3`, `dsv3` | `deepseek/deepseek-v3.2-exp` | DeepSeek V3.2 with strong reasoning (164K context) |
+| `qwen`, `qwen3.5` | `qwen/qwen3.5-397b-a17b` | Frontier 397B MoE reasoning model (262K context) |
+| `minimax`, `m2.5` | `minimax/minimax-m2.5` | SWE-Bench 80.2%, agent-optimized (197K context) |
 
 Consult the JSON file for the full list, aliases, and capability flags. Add new entries as OpenRouter releases additional models.
 

pyproject.toml

@@ -1,8 +1,8 @@
 [project]
 name = "pal-mcp-server"
-version = "9.8.2"
+version = "9.9.0"
 description = "AI-powered MCP server with multiple model providers"
-requires-python = ">=3.9"
+requires-python = ">=3.10"
 dependencies = [
     "mcp>=1.0.0",
     "google-genai>=1.19.0",

tests/test_custom_provider.py

@@ -87,9 +87,9 @@ def test_model_alias_resolution(self):
         provider = CustomProvider(api_key="test-key", base_url="http://localhost:11434/v1")
 
         # Test that aliases resolve properly
-        # "llama" now resolves to "meta-llama/llama-3-70b" (the OpenRouter model)
-        resolved = provider._resolve_model_name("llama")
-        assert resolved == "meta-llama/llama-3-70b"
+        # "deepseek" resolves to "deepseek/deepseek-r1-0528" (the OpenRouter model)
+        resolved = provider._resolve_model_name("deepseek")
+        assert resolved == "deepseek/deepseek-r1-0528"
 
         # Test local model alias
         resolved_local = provider._resolve_model_name("local-llama")
@@ -216,8 +216,8 @@ def custom_provider_factory(api_key=None):
             custom_provider = custom_provider_factory()
             openrouter_provider = OpenRouterProvider(api_key="test-openrouter-key")
 
-            assert not custom_provider.validate_model_name("llama")
-            assert openrouter_provider.validate_model_name("llama")
+            assert not custom_provider.validate_model_name("deepseek")
+            assert openrouter_provider.validate_model_name("deepseek")
 
 
 class TestConfigureProvidersFunction: