diff --git a/README.md b/README.md
index 40552da..504a5a2 100644
--- a/README.md
+++ b/README.md
@@ -409,7 +409,7 @@ for most debugging workflows, as Claude is usually able to confidently find the
 When in doubt, you can always follow up with a new prompt and ask Claude to share its findings with another model:
 
 ```text
-Use continuation with thinkdeep, share details with o4-mini-high to find out what the best fix is for this
+Use continuation with thinkdeep, share details with o4-mini to find out what the best fix is for this
 ```
 
 **[📖 Read More](docs/tools/debug.md)** - Step-by-step investigation methodology with workflow enforcement
diff --git a/communication_simulator_test.py b/communication_simulator_test.py
index 6bd1a07..e471b33 100644
--- a/communication_simulator_test.py
+++ b/communication_simulator_test.py
@@ -80,7 +80,12 @@ class CommunicationSimulator:
     """Simulates real-world Claude CLI communication with MCP Gemini server"""
 
     def __init__(
-        self, verbose: bool = False, keep_logs: bool = False, selected_tests: list[str] = None, setup: bool = False, quick_mode: bool = False
+        self,
+        verbose: bool = False,
+        keep_logs: bool = False,
+        selected_tests: list[str] = None,
+        setup: bool = False,
+        quick_mode: bool = False,
     ):
         self.verbose = verbose
         self.keep_logs = keep_logs
@@ -104,12 +109,12 @@ class CommunicationSimulator:
         # Define quick mode tests (essential tests for time-limited testing)
         # Focus on tests that work with current tool configurations
         self.quick_mode_tests = [
-            "cross_tool_continuation",        # Cross-tool conversation memory
-            "basic_conversation",             # Basic chat functionality
-            "content_validation",             # Content validation and deduplication
-            "model_thinking_config",          # Flash/flashlite model testing
-            "o3_model_selection",             # O3 model selection testing
-            "per_tool_deduplication"          # File deduplication for individual tools
+            "cross_tool_continuation",  # Cross-tool conversation memory
+            "basic_conversation",  # Basic chat functionality
+            "content_validation",  # Content validation and deduplication
+            "model_thinking_config",  # Flash/flashlite model testing
+            "o3_model_selection",  # O3 model selection testing
+            "per_tool_deduplication",  # File deduplication for individual tools
         ]
 
         # If quick mode is enabled, override selected_tests
@@ -444,7 +449,9 @@ def parse_arguments():
     parser.add_argument("--tests", "-t", nargs="+", help="Specific tests to run (space-separated)")
     parser.add_argument("--list-tests", action="store_true", help="List available tests and exit")
     parser.add_argument("--individual", "-i", help="Run a single test individually")
-    parser.add_argument("--quick", "-q", action="store_true", help="Run quick test mode (6 essential tests for time-limited testing)")
+    parser.add_argument(
+        "--quick", "-q", action="store_true", help="Run quick test mode (6 essential tests for time-limited testing)"
+    )
     parser.add_argument(
         "--setup", action="store_true", help="Force setup standalone server environment using run-server.sh"
     )
@@ -522,7 +529,11 @@ def main():
 
     # Initialize simulator consistently for all use cases
     simulator = CommunicationSimulator(
-        verbose=args.verbose, keep_logs=args.keep_logs, selected_tests=args.tests, setup=args.setup, quick_mode=args.quick
+        verbose=args.verbose,
+        keep_logs=args.keep_logs,
+        selected_tests=args.tests,
+        setup=args.setup,
+        quick_mode=args.quick,
     )
 
     # Determine execution mode and run
diff --git a/conf/custom_models.json b/conf/custom_models.json
index 2a3bcf3..f794d00 100644
--- a/conf/custom_models.json
+++ b/conf/custom_models.json
@@ -22,6 +22,7 @@
       "model_name": "The model identifier - OpenRouter format (e.g., 'anthropic/claude-opus-4') or custom model name (e.g., 'llama3.2')",
       "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 (currently none do via OpenRouter or custom APIs)",
       "supports_json_mode": "Whether the model can guarantee valid JSON output",
       "supports_function_calling": "Whether the model supports function/tool calling",
@@ -36,6 +37,7 @@
       "model_name": "my-local-model",
       "aliases": ["shortname", "nickname", "abbrev"],
       "context_window": 128000,
+      "max_output_tokens": 32768,
       "supports_extended_thinking": false,
       "supports_json_mode": true,
       "supports_function_calling": true,
@@ -52,6 +54,7 @@
       "model_name": "anthropic/claude-opus-4",
       "aliases": ["opus", "claude-opus", "claude4-opus", "claude-4-opus"],
       "context_window": 200000,
+      "max_output_tokens": 64000,
       "supports_extended_thinking": false,
       "supports_json_mode": false,
       "supports_function_calling": false,
@@ -63,6 +66,7 @@
       "model_name": "anthropic/claude-sonnet-4",
       "aliases": ["sonnet", "claude-sonnet", "claude4-sonnet", "claude-4-sonnet", "claude"],
       "context_window": 200000,
+      "max_output_tokens": 64000,
       "supports_extended_thinking": false,
       "supports_json_mode": false,
       "supports_function_calling": false,
@@ -74,6 +78,7 @@
       "model_name": "anthropic/claude-3.5-haiku",
       "aliases": ["haiku", "claude-haiku", "claude3-haiku", "claude-3-haiku"],
       "context_window": 200000,
+      "max_output_tokens": 64000,
       "supports_extended_thinking": false,
       "supports_json_mode": false,
       "supports_function_calling": false,
@@ -85,6 +90,7 @@
       "model_name": "google/gemini-2.5-pro",
       "aliases": ["pro","gemini-pro", "gemini", "pro-openrouter"],
       "context_window": 1048576,
+      "max_output_tokens": 65536,
       "supports_extended_thinking": false,
       "supports_json_mode": true,
       "supports_function_calling": false,
@@ -96,6 +102,7 @@
       "model_name": "google/gemini-2.5-flash",
       "aliases": ["flash","gemini-flash", "flash-openrouter", "flash-2.5"],
       "context_window": 1048576,
+      "max_output_tokens": 65536,
       "supports_extended_thinking": false,
       "supports_json_mode": true,
       "supports_function_calling": false,
@@ -107,6 +114,7 @@
       "model_name": "mistralai/mistral-large-2411",
       "aliases": ["mistral-large", "mistral"],
       "context_window": 128000,
+      "max_output_tokens": 32000,
       "supports_extended_thinking": false,
       "supports_json_mode": true,
       "supports_function_calling": true,
@@ -118,6 +126,7 @@
       "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,
@@ -129,6 +138,7 @@
       "model_name": "deepseek/deepseek-r1-0528",
       "aliases": ["deepseek-r1", "deepseek", "r1", "deepseek-thinking"],
       "context_window": 65536,
+      "max_output_tokens": 32768,
       "supports_extended_thinking": true,
       "supports_json_mode": true,
       "supports_function_calling": false,
@@ -140,6 +150,7 @@
       "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,
@@ -151,6 +162,7 @@
       "model_name": "openai/o3",
       "aliases": ["o3"],
       "context_window": 200000,
+      "max_output_tokens": 100000,
       "supports_extended_thinking": false,
       "supports_json_mode": true,
       "supports_function_calling": true,
@@ -164,6 +176,7 @@
       "model_name": "openai/o3-mini",
       "aliases": ["o3-mini", "o3mini"],
       "context_window": 200000,
+      "max_output_tokens": 100000,
       "supports_extended_thinking": false,
       "supports_json_mode": true,
       "supports_function_calling": true,
@@ -177,6 +190,7 @@
       "model_name": "openai/o3-mini-high",
       "aliases": ["o3-mini-high", "o3mini-high"],
       "context_window": 200000,
+      "max_output_tokens": 100000,
       "supports_extended_thinking": false,
       "supports_json_mode": true,
       "supports_function_calling": true,
@@ -190,6 +204,7 @@
       "model_name": "openai/o3-pro",
       "aliases": ["o3-pro", "o3pro"],
       "context_window": 200000,
+      "max_output_tokens": 100000,
       "supports_extended_thinking": false,
       "supports_json_mode": true,
       "supports_function_calling": true,
@@ -203,6 +218,7 @@
       "model_name": "openai/o4-mini",
       "aliases": ["o4-mini", "o4mini"],
       "context_window": 200000,
+      "max_output_tokens": 100000,
       "supports_extended_thinking": false,
       "supports_json_mode": true,
       "supports_function_calling": true,
@@ -212,23 +228,11 @@
       "temperature_constraint": "fixed",
       "description": "OpenAI's o4-mini model - optimized for shorter contexts with rapid reasoning and vision"
     },
-    {
-      "model_name": "openai/o4-mini-high",
-      "aliases": ["o4-mini-high", "o4mini-high", "o4minihigh", "o4minihi"],
-      "context_window": 200000,
-      "supports_extended_thinking": false,
-      "supports_json_mode": true,
-      "supports_function_calling": true,
-      "supports_images": true,
-      "max_image_size_mb": 20.0,
-      "supports_temperature": false,
-      "temperature_constraint": "fixed",
-      "description": "OpenAI's o4-mini with high reasoning effort - enhanced for complex tasks with vision"
-    },
     {
       "model_name": "llama3.2",
       "aliases": ["local-llama", "local", "llama3.2", "ollama-llama"],
       "context_window": 128000,
+      "max_output_tokens": 64000,
       "supports_extended_thinking": false,
       "supports_json_mode": false,
       "supports_function_calling": false,
diff --git a/docs/advanced-usage.md b/docs/advanced-usage.md
index 65dc7f3..9383354 100644
--- a/docs/advanced-usage.md
+++ b/docs/advanced-usage.md
@@ -38,7 +38,6 @@ Regardless of your default configuration, you can specify models per request:
 | **`o3`** | OpenAI | 200K tokens | Strong logical reasoning | Debugging logic errors, systematic analysis |
 | **`o3-mini`** | OpenAI | 200K tokens | Balanced speed/quality | Moderate complexity tasks |
 | **`o4-mini`** | OpenAI | 200K tokens | Latest reasoning model | Optimized for shorter contexts |
-| **`o4-mini-high`** | OpenAI | 200K tokens | Enhanced reasoning | Complex tasks requiring deeper analysis |
 | **`gpt4.1`** | OpenAI | 1M tokens | Latest GPT-4 with extended context | Large codebase analysis, comprehensive reviews |
 | **`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 |
@@ -69,7 +68,7 @@ OPENAI_ALLOWED_MODELS=o4-mini
 
 # High-performance: Quality over cost
 GOOGLE_ALLOWED_MODELS=pro
-OPENAI_ALLOWED_MODELS=o3,o4-mini-high
+OPENAI_ALLOWED_MODELS=o3,o4-mini
 ```
 
 **Important Notes:**
@@ -144,7 +143,7 @@ All tools that work with files support **both individual files and entire direct
 **`analyze`** - Analyze files or directories
 - `files`: List of file paths or directories (required)
 - `question`: What to analyze (required)  
-- `model`: auto|pro|flash|o3|o3-mini|o4-mini|o4-mini-high|gpt4.1 (default: server default)
+- `model`: auto|pro|flash|o3|o3-mini|o4-mini|gpt4.1 (default: server default)
 - `analysis_type`: architecture|performance|security|quality|general
 - `output_format`: summary|detailed|actionable
 - `thinking_mode`: minimal|low|medium|high|max (default: medium, Gemini only)
@@ -159,7 +158,7 @@ All tools that work with files support **both individual files and entire direct
 
 **`codereview`** - Review code files or directories
 - `files`: List of file paths or directories (required)
-- `model`: auto|pro|flash|o3|o3-mini|o4-mini|o4-mini-high|gpt4.1 (default: server default)
+- `model`: auto|pro|flash|o3|o3-mini|o4-mini|gpt4.1 (default: server default)
 - `review_type`: full|security|performance|quick
 - `focus_on`: Specific aspects to focus on
 - `standards`: Coding standards to enforce
@@ -175,7 +174,7 @@ All tools that work with files support **both individual files and entire direct
 
 **`debug`** - Debug with file context
 - `error_description`: Description of the issue (required)
-- `model`: auto|pro|flash|o3|o3-mini|o4-mini|o4-mini-high|gpt4.1 (default: server default)
+- `model`: auto|pro|flash|o3|o3-mini|o4-mini|gpt4.1 (default: server default)
 - `error_context`: Stack trace or logs
 - `files`: Files or directories related to the issue
 - `runtime_info`: Environment details
@@ -191,7 +190,7 @@ All tools that work with files support **both individual files and entire direct
 
 **`thinkdeep`** - Extended analysis with file context
 - `current_analysis`: Your current thinking (required)
-- `model`: auto|pro|flash|o3|o3-mini|o4-mini|o4-mini-high|gpt4.1 (default: server default)
+- `model`: auto|pro|flash|o3|o3-mini|o4-mini|gpt4.1 (default: server default)
 - `problem_context`: Additional context
 - `focus_areas`: Specific aspects to focus on
 - `files`: Files or directories for context
@@ -207,7 +206,7 @@ All tools that work with files support **both individual files and entire direct
 **`testgen`** - Comprehensive test generation with edge case coverage
 - `files`: Code files or directories to generate tests for (required)
 - `prompt`: Description of what to test, testing objectives, and scope (required)
-- `model`: auto|pro|flash|o3|o3-mini|o4-mini|o4-mini-high|gpt4.1 (default: server default)
+- `model`: auto|pro|flash|o3|o3-mini|o4-mini|gpt4.1 (default: server default)
 - `test_examples`: Optional existing test files as style/pattern reference
 - `thinking_mode`: minimal|low|medium|high|max (default: medium, Gemini only)
 
@@ -222,7 +221,7 @@ All tools that work with files support **both individual files and entire direct
 - `files`: Code files or directories to analyze for refactoring opportunities (required)
 - `prompt`: Description of refactoring goals, context, and specific areas of focus (required)
 - `refactor_type`: codesmells|decompose|modernize|organization (required)
-- `model`: auto|pro|flash|o3|o3-mini|o4-mini|o4-mini-high|gpt4.1 (default: server default)
+- `model`: auto|pro|flash|o3|o3-mini|o4-mini|gpt4.1 (default: server default)
 - `focus_areas`: Specific areas to focus on (e.g., 'performance', 'readability', 'maintainability', 'security')
 - `style_guide_examples`: Optional existing code files to use as style/pattern reference
 - `thinking_mode`: minimal|low|medium|high|max (default: medium, Gemini only)
diff --git a/docs/configuration.md b/docs/configuration.md
index 8107cc4..473b6de 100644
--- a/docs/configuration.md
+++ b/docs/configuration.md
@@ -63,7 +63,7 @@ CUSTOM_MODEL_NAME=llama3.2                          # Default model
 
 **Default Model Selection:**
 ```env
-# Options: 'auto', 'pro', 'flash', 'o3', 'o3-mini', 'o4-mini', 'o4-mini-high', etc.
+# Options: 'auto', 'pro', 'flash', 'o3', 'o3-mini', 'o4-mini', etc.
 DEFAULT_MODEL=auto  # Claude picks best model for each task (recommended)
 ```
 
@@ -74,7 +74,6 @@ DEFAULT_MODEL=auto  # Claude picks best model for each task (recommended)
 - **`o3`**: Strong logical reasoning (200K context)
 - **`o3-mini`**: Balanced speed/quality (200K context)
 - **`o4-mini`**: Latest reasoning model, optimized for shorter contexts
-- **`o4-mini-high`**: Enhanced O4 with higher reasoning effort
 - **`grok`**: GROK-3 advanced reasoning (131K context)
 - **Custom models**: via OpenRouter or local APIs
 
@@ -120,7 +119,6 @@ OPENROUTER_ALLOWED_MODELS=opus,sonnet,mistral
 - `o3` (200K context, high reasoning)
 - `o3-mini` (200K context, balanced)
 - `o4-mini` (200K context, latest balanced)
-- `o4-mini-high` (200K context, enhanced reasoning)
 - `mini` (shorthand for o4-mini)
 
 **Gemini Models:**
diff --git a/docs/tools/analyze.md b/docs/tools/analyze.md
index 379b20d..618a0be 100644
--- a/docs/tools/analyze.md
+++ b/docs/tools/analyze.md
@@ -65,7 +65,7 @@ This workflow ensures methodical analysis before expert insights, resulting in d
 
 **Initial Configuration (used in step 1):**
 - `prompt`: What to analyze or look for (required)
-- `model`: auto|pro|flash|o3|o3-mini|o4-mini|o4-mini-high|gpt4.1 (default: server default)
+- `model`: auto|pro|flash|o3|o3-mini|o4-mini|gpt4.1 (default: server default)
 - `analysis_type`: architecture|performance|security|quality|general (default: general)
 - `output_format`: summary|detailed|actionable (default: detailed)
 - `temperature`: Temperature for analysis (0-1, default 0.2)
diff --git a/docs/tools/chat.md b/docs/tools/chat.md
index 1c2b507..b7557eb 100644
--- a/docs/tools/chat.md
+++ b/docs/tools/chat.md
@@ -33,7 +33,7 @@ and then debate with the other models to give me a final verdict
 ## Tool Parameters
 
 - `prompt`: Your question or discussion topic (required)
-- `model`: auto|pro|flash|o3|o3-mini|o4-mini|o4-mini-high|gpt4.1 (default: server default)
+- `model`: auto|pro|flash|o3|o3-mini|o4-mini|gpt4.1 (default: server default)
 - `files`: Optional files for context (absolute paths)
 - `images`: Optional images for visual context (absolute paths)
 - `temperature`: Response creativity (0-1, default 0.5)
diff --git a/docs/tools/codereview.md b/docs/tools/codereview.md
index 9ba650c..9037cc2 100644
--- a/docs/tools/codereview.md
+++ b/docs/tools/codereview.md
@@ -80,7 +80,7 @@ The above prompt will simultaneously run two separate `codereview` tools with tw
 
 **Initial Review Configuration (used in step 1):**
 - `prompt`: User's summary of what the code does, expected behavior, constraints, and review objectives (required)
-- `model`: auto|pro|flash|o3|o3-mini|o4-mini|o4-mini-high|gpt4.1 (default: server default)
+- `model`: auto|pro|flash|o3|o3-mini|o4-mini|gpt4.1 (default: server default)
 - `review_type`: full|security|performance|quick (default: full)
 - `focus_on`: Specific aspects to focus on (e.g., "security vulnerabilities", "performance bottlenecks")
 - `standards`: Coding standards to enforce (e.g., "PEP8", "ESLint", "Google Style Guide")
diff --git a/docs/tools/debug.md b/docs/tools/debug.md
index 7efc454..6e7f20d 100644
--- a/docs/tools/debug.md
+++ b/docs/tools/debug.md
@@ -73,7 +73,7 @@ This structured approach ensures Claude performs methodical groundwork before ex
 - `images`: Visual debugging materials (error screenshots, logs, etc.)
 
 **Model Selection:**
-- `model`: auto|pro|flash|o3|o3-mini|o4-mini|o4-mini-high (default: server default)
+- `model`: auto|pro|flash|o3|o3-mini|o4-mini (default: server default)
 - `thinking_mode`: minimal|low|medium|high|max (default: medium, Gemini only)
 - `use_websearch`: Enable web search for documentation and solutions (default: true)
 - `use_assistant_model`: Whether to use expert analysis phase (default: true, set to false to use Claude only)
diff --git a/docs/tools/precommit.md b/docs/tools/precommit.md
index a218bd4..d70c1ab 100644
--- a/docs/tools/precommit.md
+++ b/docs/tools/precommit.md
@@ -135,7 +135,7 @@ Use zen and perform a thorough precommit ensuring there aren't any new regressio
 **Initial Configuration (used in step 1):**
 - `path`: Starting directory to search for repos (default: current directory, absolute path required)
 - `prompt`: The original user request description for the changes (required for context)
-- `model`: auto|pro|flash|o3|o3-mini|o4-mini|o4-mini-high|gpt4.1 (default: server default)
+- `model`: auto|pro|flash|o3|o3-mini|o4-mini|gpt4.1 (default: server default)
 - `compare_to`: Compare against a branch/tag instead of local changes (optional)
 - `severity_filter`: critical|high|medium|low|all (default: all)
 - `include_staged`: Include staged changes in the review (default: true)
diff --git a/docs/tools/refactor.md b/docs/tools/refactor.md
index 8314a4e..6407a4a 100644
--- a/docs/tools/refactor.md
+++ b/docs/tools/refactor.md
@@ -103,7 +103,7 @@ This results in Claude first performing its own expert analysis, encouraging it
 **Initial Configuration (used in step 1):**
 - `prompt`: Description of refactoring goals, context, and specific areas of focus (required)
 - `refactor_type`: codesmells|decompose|modernize|organization (default: codesmells)
-- `model`: auto|pro|flash|o3|o3-mini|o4-mini|o4-mini-high|gpt4.1 (default: server default)
+- `model`: auto|pro|flash|o3|o3-mini|o4-mini|gpt4.1 (default: server default)
 - `focus_areas`: Specific areas to focus on (e.g., 'performance', 'readability', 'maintainability', 'security')
 - `style_guide_examples`: Optional existing code files to use as style/pattern reference (absolute paths)
 - `thinking_mode`: minimal|low|medium|high|max (default: medium, Gemini only)
diff --git a/docs/tools/secaudit.md b/docs/tools/secaudit.md
index 36c4b8f..280452f 100644
--- a/docs/tools/secaudit.md
+++ b/docs/tools/secaudit.md
@@ -86,7 +86,7 @@ security remediation plan using planner
 - `images`: Architecture diagrams, security documentation, or visual references
 
 **Initial Security Configuration (used in step 1):**
-- `model`: auto|pro|flash|o3|o3-mini|o4-mini|o4-mini-high|gpt4.1 (default: server default)
+- `model`: auto|pro|flash|o3|o3-mini|o4-mini|gpt4.1 (default: server default)
 - `security_scope`: Application context, technology stack, and security boundary definition (required)
 - `threat_level`: low|medium|high|critical (default: medium) - determines assessment depth and urgency
 - `compliance_requirements`: List of compliance frameworks to assess against (e.g., ["PCI DSS", "SOC2"])
diff --git a/docs/tools/testgen.md b/docs/tools/testgen.md
index e19d042..0d74a98 100644
--- a/docs/tools/testgen.md
+++ b/docs/tools/testgen.md
@@ -70,7 +70,7 @@ Test generation excels with extended reasoning models like Gemini Pro or O3, whi
 
 **Initial Configuration (used in step 1):**
 - `prompt`: Description of what to test, testing objectives, and specific scope/focus areas (required)
-- `model`: auto|pro|flash|o3|o3-mini|o4-mini|o4-mini-high|gpt4.1 (default: server default)
+- `model`: auto|pro|flash|o3|o3-mini|o4-mini|gpt4.1 (default: server default)
 - `test_examples`: Optional existing test files or directories to use as style/pattern reference (absolute paths)
 - `thinking_mode`: minimal|low|medium|high|max (default: medium, Gemini only)
 - `use_assistant_model`: Whether to use expert test generation phase (default: true, set to false to use Claude only)
diff --git a/docs/tools/thinkdeep.md b/docs/tools/thinkdeep.md
index 5180a8b..26d5322 100644
--- a/docs/tools/thinkdeep.md
+++ b/docs/tools/thinkdeep.md
@@ -30,7 +30,7 @@ with the best architecture for my project
 ## Tool Parameters
 
 - `prompt`: Your current thinking/analysis to extend and validate (required)
-- `model`: auto|pro|flash|o3|o3-mini|o4-mini|o4-mini-high|gpt4.1 (default: server default)
+- `model`: auto|pro|flash|o3|o3-mini|o4-mini|gpt4.1 (default: server default)
 - `problem_context`: Additional context about the problem or goal
 - `focus_areas`: Specific aspects to focus on (architecture, performance, security, etc.)
 - `files`: Optional file paths or directories for additional context (absolute paths)
diff --git a/providers/base.py b/providers/base.py
index 06f60fe..aff8705 100644
--- a/providers/base.py
+++ b/providers/base.py
@@ -132,6 +132,7 @@ class ModelCapabilities:
     model_name: str
     friendly_name: str  # Human-friendly name like "Gemini" or "OpenAI"
     context_window: int  # Total context window size in tokens
+    max_output_tokens: int  # Maximum output tokens per request
     supports_extended_thinking: bool = False
     supports_system_prompts: bool = True
     supports_streaming: bool = True
diff --git a/providers/custom.py b/providers/custom.py
index 021bba5..d32d494 100644
--- a/providers/custom.py
+++ b/providers/custom.py
@@ -158,6 +158,7 @@ class CustomProvider(OpenAICompatibleProvider):
                 model_name=resolved_name,
                 friendly_name=f"{self.FRIENDLY_NAME} ({resolved_name})",
                 context_window=32_768,  # Conservative default
+                max_output_tokens=32_768,  # Conservative default max output
                 supports_extended_thinking=False,  # Most custom models don't support this
                 supports_system_prompts=True,
                 supports_streaming=True,
@@ -187,7 +188,7 @@ class CustomProvider(OpenAICompatibleProvider):
         Returns:
             True if model is intended for custom/local endpoint
         """
-        logging.debug(f"Custom provider validating model: '{model_name}'")
+        # logging.debug(f"Custom provider validating model: '{model_name}'")
 
         # Try to resolve through registry first
         config = self._registry.resolve(model_name)
@@ -195,12 +196,12 @@ class CustomProvider(OpenAICompatibleProvider):
             model_id = config.model_name
             # Use explicit is_custom flag for clean validation
             if config.is_custom:
-                logging.debug(f"Model '{model_name}' -> '{model_id}' validated via registry (custom model)")
+                logging.debug(f"... [Custom] Model '{model_name}' -> '{model_id}' validated via registry")
                 return True
             else:
                 # This is a cloud/OpenRouter model - CustomProvider should NOT handle these
                 # Let OpenRouter provider handle them instead
-                logging.debug(f"Model '{model_name}' -> '{model_id}' rejected (cloud model, defer to OpenRouter)")
+                # logging.debug(f"... [Custom] Model '{model_name}' -> '{model_id}' not custom (defer to OpenRouter)")
                 return False
 
         # Handle version tags for unknown models (e.g., "my-model:latest")
diff --git a/providers/dial.py b/providers/dial.py
index f019415..e0c4a29 100644
--- a/providers/dial.py
+++ b/providers/dial.py
@@ -37,6 +37,7 @@ class DIALModelProvider(OpenAICompatibleProvider):
             model_name="o3-2025-04-16",
             friendly_name="DIAL (O3)",
             context_window=200_000,
+            max_output_tokens=100_000,
             supports_extended_thinking=False,
             supports_system_prompts=True,
             supports_streaming=True,
@@ -54,6 +55,7 @@ class DIALModelProvider(OpenAICompatibleProvider):
             model_name="o4-mini-2025-04-16",
             friendly_name="DIAL (O4-mini)",
             context_window=200_000,
+            max_output_tokens=100_000,
             supports_extended_thinking=False,
             supports_system_prompts=True,
             supports_streaming=True,
@@ -71,6 +73,7 @@ class DIALModelProvider(OpenAICompatibleProvider):
             model_name="anthropic.claude-sonnet-4-20250514-v1:0",
             friendly_name="DIAL (Sonnet 4)",
             context_window=200_000,
+            max_output_tokens=64_000,
             supports_extended_thinking=False,
             supports_system_prompts=True,
             supports_streaming=True,
@@ -88,6 +91,7 @@ class DIALModelProvider(OpenAICompatibleProvider):
             model_name="anthropic.claude-sonnet-4-20250514-v1:0-with-thinking",
             friendly_name="DIAL (Sonnet 4 Thinking)",
             context_window=200_000,
+            max_output_tokens=64_000,
             supports_extended_thinking=True,  # Thinking mode variant
             supports_system_prompts=True,
             supports_streaming=True,
@@ -105,6 +109,7 @@ class DIALModelProvider(OpenAICompatibleProvider):
             model_name="anthropic.claude-opus-4-20250514-v1:0",
             friendly_name="DIAL (Opus 4)",
             context_window=200_000,
+            max_output_tokens=64_000,
             supports_extended_thinking=False,
             supports_system_prompts=True,
             supports_streaming=True,
@@ -122,6 +127,7 @@ class DIALModelProvider(OpenAICompatibleProvider):
             model_name="anthropic.claude-opus-4-20250514-v1:0-with-thinking",
             friendly_name="DIAL (Opus 4 Thinking)",
             context_window=200_000,
+            max_output_tokens=64_000,
             supports_extended_thinking=True,  # Thinking mode variant
             supports_system_prompts=True,
             supports_streaming=True,
@@ -139,6 +145,7 @@ class DIALModelProvider(OpenAICompatibleProvider):
             model_name="gemini-2.5-pro-preview-03-25-google-search",
             friendly_name="DIAL (Gemini 2.5 Pro Search)",
             context_window=1_000_000,
+            max_output_tokens=65_536,
             supports_extended_thinking=False,  # DIAL doesn't expose thinking mode
             supports_system_prompts=True,
             supports_streaming=True,
@@ -156,6 +163,7 @@ class DIALModelProvider(OpenAICompatibleProvider):
             model_name="gemini-2.5-pro-preview-05-06",
             friendly_name="DIAL (Gemini 2.5 Pro)",
             context_window=1_000_000,
+            max_output_tokens=65_536,
             supports_extended_thinking=False,
             supports_system_prompts=True,
             supports_streaming=True,
@@ -173,6 +181,7 @@ class DIALModelProvider(OpenAICompatibleProvider):
             model_name="gemini-2.5-flash-preview-05-20",
             friendly_name="DIAL (Gemini Flash 2.5)",
             context_window=1_000_000,
+            max_output_tokens=65_536,
             supports_extended_thinking=False,
             supports_system_prompts=True,
             supports_streaming=True,
diff --git a/providers/gemini.py b/providers/gemini.py
index 1118699..51916b0 100644
--- a/providers/gemini.py
+++ b/providers/gemini.py
@@ -24,6 +24,7 @@ class GeminiModelProvider(ModelProvider):
             model_name="gemini-2.0-flash",
             friendly_name="Gemini (Flash 2.0)",
             context_window=1_048_576,  # 1M tokens
+            max_output_tokens=65_536,
             supports_extended_thinking=True,  # Experimental thinking mode
             supports_system_prompts=True,
             supports_streaming=True,
@@ -42,6 +43,7 @@ class GeminiModelProvider(ModelProvider):
             model_name="gemini-2.0-flash-lite",
             friendly_name="Gemin (Flash Lite 2.0)",
             context_window=1_048_576,  # 1M tokens
+            max_output_tokens=65_536,
             supports_extended_thinking=False,  # Not supported per user request
             supports_system_prompts=True,
             supports_streaming=True,
@@ -59,6 +61,7 @@ class GeminiModelProvider(ModelProvider):
             model_name="gemini-2.5-flash",
             friendly_name="Gemini (Flash 2.5)",
             context_window=1_048_576,  # 1M tokens
+            max_output_tokens=65_536,
             supports_extended_thinking=True,
             supports_system_prompts=True,
             supports_streaming=True,
@@ -77,6 +80,7 @@ class GeminiModelProvider(ModelProvider):
             model_name="gemini-2.5-pro",
             friendly_name="Gemini (Pro 2.5)",
             context_window=1_048_576,  # 1M tokens
+            max_output_tokens=65_536,
             supports_extended_thinking=True,
             supports_system_prompts=True,
             supports_streaming=True,
diff --git a/providers/openai_compatible.py b/providers/openai_compatible.py
index fec4484..17ce60d 100644
--- a/providers/openai_compatible.py
+++ b/providers/openai_compatible.py
@@ -687,7 +687,6 @@ class OpenAICompatibleProvider(ModelProvider):
             "o3-mini",
             "o3-pro",
             "o4-mini",
-            "o4-mini-high",
             # Note: Claude models would be handled by a separate provider
         }
         supports = model_name.lower() in vision_models
diff --git a/providers/openai_provider.py b/providers/openai_provider.py
index e065ee1..d977869 100644
--- a/providers/openai_provider.py
+++ b/providers/openai_provider.py
@@ -24,6 +24,7 @@ class OpenAIModelProvider(OpenAICompatibleProvider):
             model_name="o3",
             friendly_name="OpenAI (O3)",
             context_window=200_000,  # 200K tokens
+            max_output_tokens=65536,  # 64K max output tokens
             supports_extended_thinking=False,
             supports_system_prompts=True,
             supports_streaming=True,
@@ -41,6 +42,7 @@ class OpenAIModelProvider(OpenAICompatibleProvider):
             model_name="o3-mini",
             friendly_name="OpenAI (O3-mini)",
             context_window=200_000,  # 200K tokens
+            max_output_tokens=65536,  # 64K max output tokens
             supports_extended_thinking=False,
             supports_system_prompts=True,
             supports_streaming=True,
@@ -58,6 +60,7 @@ class OpenAIModelProvider(OpenAICompatibleProvider):
             model_name="o3-pro-2025-06-10",
             friendly_name="OpenAI (O3-Pro)",
             context_window=200_000,  # 200K tokens
+            max_output_tokens=65536,  # 64K max output tokens
             supports_extended_thinking=False,
             supports_system_prompts=True,
             supports_streaming=True,
@@ -75,6 +78,7 @@ class OpenAIModelProvider(OpenAICompatibleProvider):
             model_name="o4-mini",
             friendly_name="OpenAI (O4-mini)",
             context_window=200_000,  # 200K tokens
+            max_output_tokens=65536,  # 64K max output tokens
             supports_extended_thinking=False,
             supports_system_prompts=True,
             supports_streaming=True,
@@ -85,30 +89,14 @@ class OpenAIModelProvider(OpenAICompatibleProvider):
             supports_temperature=False,  # O4 models don't accept temperature parameter
             temperature_constraint=create_temperature_constraint("fixed"),
             description="Latest reasoning model (200K context) - Optimized for shorter contexts, rapid reasoning",
-            aliases=["mini", "o4mini"],
-        ),
-        "o4-mini-high": ModelCapabilities(
-            provider=ProviderType.OPENAI,
-            model_name="o4-mini-high",
-            friendly_name="OpenAI (O4-mini-high)",
-            context_window=200_000,  # 200K tokens
-            supports_extended_thinking=False,
-            supports_system_prompts=True,
-            supports_streaming=True,
-            supports_function_calling=True,
-            supports_json_mode=True,
-            supports_images=True,  # O4 models support vision
-            max_image_size_mb=20.0,  # 20MB per OpenAI docs
-            supports_temperature=False,  # O4 models don't accept temperature parameter
-            temperature_constraint=create_temperature_constraint("fixed"),
-            description="Enhanced O4 mini (200K context) - Higher reasoning effort for complex tasks",
-            aliases=["o4minihigh", "o4minihi", "mini-high"],
+            aliases=["mini", "o4mini", "o4-mini"],
         ),
         "gpt-4.1-2025-04-14": ModelCapabilities(
             provider=ProviderType.OPENAI,
             model_name="gpt-4.1-2025-04-14",
             friendly_name="OpenAI (GPT 4.1)",
             context_window=1_000_000,  # 1M tokens
+            max_output_tokens=32_768,
             supports_extended_thinking=False,
             supports_system_prompts=True,
             supports_streaming=True,
diff --git a/providers/openrouter.py b/providers/openrouter.py
index 3d90238..18d3d5e 100644
--- a/providers/openrouter.py
+++ b/providers/openrouter.py
@@ -101,6 +101,7 @@ class OpenRouterProvider(OpenAICompatibleProvider):
                 model_name=resolved_name,
                 friendly_name=self.FRIENDLY_NAME,
                 context_window=32_768,  # Conservative default context window
+                max_output_tokens=32_768,
                 supports_extended_thinking=False,
                 supports_system_prompts=True,
                 supports_streaming=True,
diff --git a/providers/registry.py b/providers/registry.py
index da7a9b5..4ab5732 100644
--- a/providers/registry.py
+++ b/providers/registry.py
@@ -24,8 +24,6 @@ class ModelProviderRegistry:
             cls._instance._providers = {}
             cls._instance._initialized_providers = {}
             logging.debug(f"REGISTRY: Created instance {cls._instance}")
-        else:
-            logging.debug(f"REGISTRY: Returning existing instance {cls._instance}")
         return cls._instance
 
     @classmethod
diff --git a/providers/xai.py b/providers/xai.py
index 2b6fd04..dcb14a1 100644
--- a/providers/xai.py
+++ b/providers/xai.py
@@ -26,6 +26,7 @@ class XAIModelProvider(OpenAICompatibleProvider):
             model_name="grok-3",
             friendly_name="X.AI (Grok 3)",
             context_window=131_072,  # 131K tokens
+            max_output_tokens=131072,
             supports_extended_thinking=False,
             supports_system_prompts=True,
             supports_streaming=True,
@@ -43,6 +44,7 @@ class XAIModelProvider(OpenAICompatibleProvider):
             model_name="grok-3-fast",
             friendly_name="X.AI (Grok 3 Fast)",
             context_window=131_072,  # 131K tokens
+            max_output_tokens=131072,
             supports_extended_thinking=False,
             supports_system_prompts=True,
             supports_streaming=True,
diff --git a/tests/mock_helpers.py b/tests/mock_helpers.py
index eb283b6..1122af1 100644
--- a/tests/mock_helpers.py
+++ b/tests/mock_helpers.py
@@ -15,6 +15,7 @@ def create_mock_provider(model_name="gemini-2.5-flash", context_window=1_048_576
         model_name=model_name,
         friendly_name="Gemini",
         context_window=context_window,
+        max_output_tokens=8192,
         supports_extended_thinking=False,
         supports_system_prompts=True,
         supports_streaming=True,
diff --git a/tests/test_alias_target_restrictions.py b/tests/test_alias_target_restrictions.py
index 7b182e6..dd36b83 100644
--- a/tests/test_alias_target_restrictions.py
+++ b/tests/test_alias_target_restrictions.py
@@ -211,7 +211,7 @@ class TestAliasTargetRestrictions:
         # Verify the polymorphic method was called
         mock_provider.list_all_known_models.assert_called_once()
 
-    @patch.dict(os.environ, {"OPENAI_ALLOWED_MODELS": "o4-mini-high"})  # Restrict to specific model
+    @patch.dict(os.environ, {"OPENAI_ALLOWED_MODELS": "o4-mini"})  # Restrict to specific model
     def test_complex_alias_chains_handled_correctly(self):
         """Test that complex alias chains are handled correctly in restrictions."""
         # Clear cached restriction service
@@ -221,12 +221,11 @@ class TestAliasTargetRestrictions:
 
         provider = OpenAIModelProvider(api_key="test-key")
 
-        # Only o4-mini-high should be allowed
-        assert provider.validate_model_name("o4-mini-high")
+        # Only o4-mini should be allowed
+        assert provider.validate_model_name("o4-mini")
 
         # Other models should be blocked
-        assert not provider.validate_model_name("o4-mini")
-        assert not provider.validate_model_name("mini")  # This resolves to o4-mini
+        assert not provider.validate_model_name("o3")
         assert not provider.validate_model_name("o3-mini")
 
     def test_critical_regression_validation_sees_alias_targets(self):
@@ -307,7 +306,7 @@ class TestAliasTargetRestrictions:
         it appear that target-based restrictions don't work.
         """
         # Test with a made-up restriction scenario
-        with patch.dict(os.environ, {"OPENAI_ALLOWED_MODELS": "o4-mini-high,o3-mini"}):
+        with patch.dict(os.environ, {"OPENAI_ALLOWED_MODELS": "o4-mini,o3-mini"}):
             # Clear cached restriction service
             import utils.model_restrictions
 
@@ -318,7 +317,7 @@ class TestAliasTargetRestrictions:
 
             # These specific target models should be recognized as valid
             all_known = provider.list_all_known_models()
-            assert "o4-mini-high" in all_known, "Target model o4-mini-high should be known"
+            assert "o4-mini" in all_known, "Target model o4-mini should be known"
             assert "o3-mini" in all_known, "Target model o3-mini should be known"
 
             # Validation should not warn about these being unrecognized
@@ -329,11 +328,11 @@ class TestAliasTargetRestrictions:
                 # Should not warn about our allowed models being unrecognized
                 all_warnings = [str(call) for call in mock_logger.warning.call_args_list]
                 for warning in all_warnings:
-                    assert "o4-mini-high" not in warning or "not a recognized" not in warning
+                    assert "o4-mini" not in warning or "not a recognized" not in warning
                     assert "o3-mini" not in warning or "not a recognized" not in warning
 
             # The restriction should actually work
-            assert provider.validate_model_name("o4-mini-high")
+            assert provider.validate_model_name("o4-mini")
             assert provider.validate_model_name("o3-mini")
-            assert not provider.validate_model_name("o4-mini")  # not in allowed list
+            assert not provider.validate_model_name("o3-pro")  # not in allowed list
             assert not provider.validate_model_name("o3")  # not in allowed list
diff --git a/tests/test_auto_mode.py b/tests/test_auto_mode.py
index 74d8ae3..f96feb3 100644
--- a/tests/test_auto_mode.py
+++ b/tests/test_auto_mode.py
@@ -64,7 +64,7 @@ class TestAutoMode:
                         models_with_descriptions[model_name] = description
 
         # Check all expected models are present with meaningful descriptions
-        expected_models = ["flash", "pro", "o3", "o3-mini", "o3-pro", "o4-mini", "o4-mini-high"]
+        expected_models = ["flash", "pro", "o3", "o3-mini", "o3-pro", "o4-mini"]
         for model in expected_models:
             # Model should exist somewhere in the providers
             # Note: Some models might not be available if API keys aren't configured
diff --git a/tests/test_buggy_behavior_prevention.py b/tests/test_buggy_behavior_prevention.py
index e960f1f..e925e31 100644
--- a/tests/test_buggy_behavior_prevention.py
+++ b/tests/test_buggy_behavior_prevention.py
@@ -118,7 +118,7 @@ class TestBuggyBehaviorPrevention:
         provider = OpenAIModelProvider(api_key="test-key")
 
         # Simulate a scenario where admin wants to restrict specific targets
-        with patch.dict(os.environ, {"OPENAI_ALLOWED_MODELS": "o3-mini,o4-mini-high"}):
+        with patch.dict(os.environ, {"OPENAI_ALLOWED_MODELS": "o3-mini,o4-mini"}):
             # Clear cached restriction service
             import utils.model_restrictions
 
@@ -126,19 +126,21 @@ class TestBuggyBehaviorPrevention:
 
             # These should work because they're explicitly allowed
             assert provider.validate_model_name("o3-mini")
-            assert provider.validate_model_name("o4-mini-high")
+            assert provider.validate_model_name("o4-mini")
 
             # These should be blocked
-            assert not provider.validate_model_name("o4-mini")  # Not in allowed list
+            assert not provider.validate_model_name("o3-pro")  # Not in allowed list
             assert not provider.validate_model_name("o3")  # Not in allowed list
-            assert not provider.validate_model_name("mini")  # Resolves to o4-mini, not allowed
+
+            # This should be ALLOWED because it resolves to o4-mini which is in the allowed list
+            assert provider.validate_model_name("mini")  # Resolves to o4-mini, which IS allowed
 
             # Verify our list_all_known_models includes the restricted models
             all_known = provider.list_all_known_models()
             assert "o3-mini" in all_known  # Should be known (and allowed)
-            assert "o4-mini-high" in all_known  # Should be known (and allowed)
-            assert "o4-mini" in all_known  # Should be known (but blocked)
-            assert "mini" in all_known  # Should be known (but blocked)
+            assert "o4-mini" in all_known  # Should be known (and allowed)
+            assert "o3-pro" in all_known  # Should be known (but blocked)
+            assert "mini" in all_known  # Should be known (and allowed since it resolves to o4-mini)
 
     def test_demonstration_of_old_vs_new_interface(self):
         """
diff --git a/tests/test_model_enumeration.py b/tests/test_model_enumeration.py
index 548f785..0a78b17 100644
--- a/tests/test_model_enumeration.py
+++ b/tests/test_model_enumeration.py
@@ -149,7 +149,7 @@ class TestModelEnumeration:
             ("o3", False),  # OpenAI - not available without API key
             ("grok", False),  # X.AI - not available without API key
             ("gemini-2.5-flash", False),  # Full Gemini name - not available without API key
-            ("o4-mini-high", False),  # OpenAI variant - not available without API key
+            ("o4-mini", False),  # OpenAI variant - not available without API key
             ("grok-3-fast", False),  # X.AI variant - not available without API key
         ],
     )
diff --git a/tests/test_model_restrictions.py b/tests/test_model_restrictions.py
index bd34a81..6a93bd5 100644
--- a/tests/test_model_restrictions.py
+++ b/tests/test_model_restrictions.py
@@ -93,7 +93,7 @@ class TestModelRestrictionService:
         with patch.dict(os.environ, {"OPENAI_ALLOWED_MODELS": "o3-mini,o4-mini"}):
             service = ModelRestrictionService()
 
-            models = ["o3", "o3-mini", "o4-mini", "o4-mini-high"]
+            models = ["o3", "o3-mini", "o4-mini", "o3-pro"]
             filtered = service.filter_models(ProviderType.OPENAI, models)
 
             assert filtered == ["o3-mini", "o4-mini"]
@@ -573,7 +573,7 @@ class TestShorthandRestrictions:
 
         # Other models should not work
         assert not openai_provider.validate_model_name("o3")
-        assert not openai_provider.validate_model_name("o4-mini-high")
+        assert not openai_provider.validate_model_name("o3-pro")
 
     @patch.dict(
         os.environ,
diff --git a/tests/test_o3_temperature_fix_simple.py b/tests/test_o3_temperature_fix_simple.py
index da0ea60..0a27256 100644
--- a/tests/test_o3_temperature_fix_simple.py
+++ b/tests/test_o3_temperature_fix_simple.py
@@ -185,7 +185,7 @@ class TestO3TemperatureParameterFixSimple:
         provider = OpenAIModelProvider(api_key="test-key")
 
         # Test O3/O4 models that should NOT support temperature parameter
-        o3_o4_models = ["o3", "o3-mini", "o3-pro", "o4-mini", "o4-mini-high"]
+        o3_o4_models = ["o3", "o3-mini", "o3-pro", "o4-mini"]
 
         for model in o3_o4_models:
             capabilities = provider.get_capabilities(model)
diff --git a/tests/test_openai_provider.py b/tests/test_openai_provider.py
index baab182..3429be9 100644
--- a/tests/test_openai_provider.py
+++ b/tests/test_openai_provider.py
@@ -47,14 +47,13 @@ class TestOpenAIProvider:
         assert provider.validate_model_name("o3-mini") is True
         assert provider.validate_model_name("o3-pro") is True
         assert provider.validate_model_name("o4-mini") is True
-        assert provider.validate_model_name("o4-mini-high") is True
+        assert provider.validate_model_name("o4-mini") is True
 
         # Test valid aliases
         assert provider.validate_model_name("mini") is True
         assert provider.validate_model_name("o3mini") is True
         assert provider.validate_model_name("o4mini") is True
-        assert provider.validate_model_name("o4minihigh") is True
-        assert provider.validate_model_name("o4minihi") is True
+        assert provider.validate_model_name("o4mini") is True
 
         # Test invalid model
         assert provider.validate_model_name("invalid-model") is False
@@ -69,15 +68,14 @@ class TestOpenAIProvider:
         assert provider._resolve_model_name("mini") == "o4-mini"
         assert provider._resolve_model_name("o3mini") == "o3-mini"
         assert provider._resolve_model_name("o4mini") == "o4-mini"
-        assert provider._resolve_model_name("o4minihigh") == "o4-mini-high"
-        assert provider._resolve_model_name("o4minihi") == "o4-mini-high"
+        assert provider._resolve_model_name("o4mini") == "o4-mini"
 
         # Test full name passthrough
         assert provider._resolve_model_name("o3") == "o3"
         assert provider._resolve_model_name("o3-mini") == "o3-mini"
         assert provider._resolve_model_name("o3-pro") == "o3-pro-2025-06-10"
         assert provider._resolve_model_name("o4-mini") == "o4-mini"
-        assert provider._resolve_model_name("o4-mini-high") == "o4-mini-high"
+        assert provider._resolve_model_name("o4-mini") == "o4-mini"
 
     def test_get_capabilities_o3(self):
         """Test getting model capabilities for O3."""
@@ -184,11 +182,11 @@ class TestOpenAIProvider:
         call_kwargs = mock_client.chat.completions.create.call_args[1]
         assert call_kwargs["model"] == "o3-mini"
 
-        # Test o4minihigh -> o4-mini-high
-        mock_response.model = "o4-mini-high"
-        provider.generate_content(prompt="Test", model_name="o4minihigh", temperature=1.0)
+        # Test o4mini -> o4-mini
+        mock_response.model = "o4-mini"
+        provider.generate_content(prompt="Test", model_name="o4mini", temperature=1.0)
         call_kwargs = mock_client.chat.completions.create.call_args[1]
-        assert call_kwargs["model"] == "o4-mini-high"
+        assert call_kwargs["model"] == "o4-mini"
 
     @patch("providers.openai_compatible.OpenAI")
     def test_generate_content_no_alias_passthrough(self, mock_openai_class):
diff --git a/tests/test_openrouter_provider.py b/tests/test_openrouter_provider.py
index 6d427ba..454f372 100644
--- a/tests/test_openrouter_provider.py
+++ b/tests/test_openrouter_provider.py
@@ -77,7 +77,7 @@ class TestOpenRouterProvider:
         assert provider._resolve_model_name("o3-mini") == "openai/o3-mini"
         assert provider._resolve_model_name("o3mini") == "openai/o3-mini"
         assert provider._resolve_model_name("o4-mini") == "openai/o4-mini"
-        assert provider._resolve_model_name("o4-mini-high") == "openai/o4-mini-high"
+        assert provider._resolve_model_name("o4-mini") == "openai/o4-mini"
         assert provider._resolve_model_name("claude") == "anthropic/claude-sonnet-4"
         assert provider._resolve_model_name("mistral") == "mistralai/mistral-large-2411"
         assert provider._resolve_model_name("deepseek") == "deepseek/deepseek-r1-0528"
diff --git a/tests/test_openrouter_registry.py b/tests/test_openrouter_registry.py
index f6ea000..6387ebe 100644
--- a/tests/test_openrouter_registry.py
+++ b/tests/test_openrouter_registry.py
@@ -24,7 +24,16 @@ class TestOpenRouterModelRegistry:
     def test_custom_config_path(self):
         """Test registry with custom config path."""
         # Create temporary config
-        config_data = {"models": [{"model_name": "test/model-1", "aliases": ["test1", "t1"], "context_window": 4096}]}
+        config_data = {
+            "models": [
+                {
+                    "model_name": "test/model-1",
+                    "aliases": ["test1", "t1"],
+                    "context_window": 4096,
+                    "max_output_tokens": 2048,
+                }
+            ]
+        }
 
         with tempfile.NamedTemporaryFile(mode="w", suffix=".json", delete=False) as f:
             json.dump(config_data, f)
@@ -42,7 +51,11 @@ class TestOpenRouterModelRegistry:
     def test_environment_variable_override(self):
         """Test OPENROUTER_MODELS_PATH environment variable."""
         # Create custom config
-        config_data = {"models": [{"model_name": "env/model", "aliases": ["envtest"], "context_window": 8192}]}
+        config_data = {
+            "models": [
+                {"model_name": "env/model", "aliases": ["envtest"], "context_window": 8192, "max_output_tokens": 4096}
+            ]
+        }
 
         with tempfile.NamedTemporaryFile(mode="w", suffix=".json", delete=False) as f:
             json.dump(config_data, f)
@@ -127,11 +140,12 @@ class TestOpenRouterModelRegistry:
         """Test that duplicate aliases are detected."""
         config_data = {
             "models": [
-                {"model_name": "test/model-1", "aliases": ["dupe"], "context_window": 4096},
+                {"model_name": "test/model-1", "aliases": ["dupe"], "context_window": 4096, "max_output_tokens": 2048},
                 {
                     "model_name": "test/model-2",
                     "aliases": ["DUPE"],  # Same alias, different case
                     "context_window": 8192,
+                    "max_output_tokens": 2048,
                 },
             ]
         }
@@ -207,6 +221,7 @@ class TestOpenRouterModelRegistry:
             friendly_name="OpenRouter (test/full-featured)",
             aliases=["full"],
             context_window=128000,
+            max_output_tokens=8192,
             supports_extended_thinking=True,
             supports_system_prompts=True,
             supports_streaming=True,
diff --git a/tests/test_providers.py b/tests/test_providers.py
index 5401bc9..036ae9b 100644
--- a/tests/test_providers.py
+++ b/tests/test_providers.py
@@ -215,9 +215,7 @@ class TestOpenAIProvider:
         assert provider.validate_model_name("o3-mini")  # Backwards compatibility
         assert provider.validate_model_name("o4-mini")
         assert provider.validate_model_name("o4mini")
-        assert provider.validate_model_name("o4-mini-high")
-        assert provider.validate_model_name("o4minihigh")
-        assert provider.validate_model_name("o4minihi")
+        assert provider.validate_model_name("o4-mini")
         assert not provider.validate_model_name("gpt-4o")
         assert not provider.validate_model_name("invalid-model")
 
@@ -229,4 +227,4 @@ class TestOpenAIProvider:
         assert not provider.supports_thinking_mode("o3mini")
         assert not provider.supports_thinking_mode("o3-mini")
         assert not provider.supports_thinking_mode("o4-mini")
-        assert not provider.supports_thinking_mode("o4-mini-high")
+        assert not provider.supports_thinking_mode("o4-mini")
diff --git a/tests/test_supported_models_aliases.py b/tests/test_supported_models_aliases.py
index 6ed899f..1eb76b5 100644
--- a/tests/test_supported_models_aliases.py
+++ b/tests/test_supported_models_aliases.py
@@ -51,15 +51,14 @@ class TestSupportedModelsAliases:
         assert "o4mini" in provider.SUPPORTED_MODELS["o4-mini"].aliases
         assert "o3mini" in provider.SUPPORTED_MODELS["o3-mini"].aliases
         assert "o3-pro" in provider.SUPPORTED_MODELS["o3-pro-2025-06-10"].aliases
-        assert "o4minihigh" in provider.SUPPORTED_MODELS["o4-mini-high"].aliases
-        assert "o4minihi" in provider.SUPPORTED_MODELS["o4-mini-high"].aliases
+        assert "o4mini" in provider.SUPPORTED_MODELS["o4-mini"].aliases
         assert "gpt4.1" in provider.SUPPORTED_MODELS["gpt-4.1-2025-04-14"].aliases
 
         # Test alias resolution
         assert provider._resolve_model_name("mini") == "o4-mini"
         assert provider._resolve_model_name("o3mini") == "o3-mini"
         assert provider._resolve_model_name("o3-pro") == "o3-pro-2025-06-10"
-        assert provider._resolve_model_name("o4minihigh") == "o4-mini-high"
+        assert provider._resolve_model_name("o4mini") == "o4-mini"
         assert provider._resolve_model_name("gpt4.1") == "gpt-4.1-2025-04-14"
 
         # Test case insensitive resolution