klaude-code 2.3.0__tar.gz → 2.4.1__tar.gz

{klaude_code-2.3.0 → klaude_code-2.4.1}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: klaude-code
-Version: 2.3.0
+Version: 2.4.1
 Summary: Minimal code agent CLI
 Requires-Dist: anthropic>=0.66.0
 Requires-Dist: chardet>=5.2.0
@@ -155,95 +155,42 @@ Open in editor:
 klaude config
 ```
 
-##### Adding Models to Built-in Providers
+##### Model Configuration
 
-You can add custom models to existing built-in providers without redefining the entire provider. Just reference the `provider_name` and add your `model_list`:
+You can add custom models to built-in providers or define new ones. Configuration is inherited from built-in providers by matching `provider_name`.
 
 ```yaml
 # ~/.klaude/klaude-config.yaml
 provider_list:
-  - provider_name: openrouter  # Reference existing built-in provider
-    model_list:
-      - model_name: seed
-        model_params:
-          model: bytedance-seed/seed-1.6  # Model ID from OpenRouter
-          context_limit: 262000
-          cost:
-            input: 0.25
-            output: 2
-```
-
-**How merging works:**
-- Your models are merged with the built-in models for that provider
-- You only need `provider_name` and `model_list` - protocol, api_key, etc. are inherited from the built-in config
-- To override a built-in model, use the same `model_name` (e.g., `sonnet` to customize the built-in sonnet)
-
-**More examples:**
-
-```yaml
-provider_list:
-  # Add multiple models to OpenRouter
+  # Add/Override models for built-in OpenRouter provider
   - provider_name: openrouter
     model_list:
       - model_name: qwen-coder
-        model_params:
-          model: qwen/qwen-2.5-coder-32b-instruct
-          context_limit: 131072
-          cost:
-            input: 0.3
-            output: 0.9
-      - model_name: llama-405b
-        model_params:
-          model: meta-llama/llama-3.1-405b-instruct
-          context_limit: 131072
-          cost:
-            input: 0.8
-            output: 0.8
-
-  # Add models to Anthropic provider
-  - provider_name: anthropic
-    model_list:
-      - model_name: haiku@ant
-        model_params:
-          model: claude-3-5-haiku-20241022
-          context_limit: 200000
-          cost:
-            input: 1.0
-            output: 5.0
-```
-
-After adding models, run `klaude list` to verify they appear in the model list.
-
-##### Overriding Provider Settings
-
-Override provider-level settings (like api_key) while keeping built-in models:
-
-```yaml
-provider_list:
-  - provider_name: anthropic
-    api_key: sk-my-custom-key  # Override the default ${ANTHROPIC_API_KEY}
-    # Built-in models (sonnet, opus) are still available
-```
-
-##### Adding New Providers
-
-For providers not in the built-in list, you must specify `protocol`:
-
-```yaml
-provider_list:
-  - provider_name: my-azure-openai
+        model_id: qwen/qwen-2.5-coder-32b-instruct
+        context_limit: 131072
+        cost: { input: 0.3, output: 0.9 }
+      - model_name: sonnet # Override built-in sonnet params
+        model_id: anthropic/claude-3.5-sonnet
+        context_limit: 200000
+
+  # Add a completely new provider
+  - provider_name: my-azure
     protocol: openai
     api_key: ${AZURE_OPENAI_KEY}
     base_url: https://my-instance.openai.azure.com/
     is_azure: true
     azure_api_version: "2024-02-15-preview"
     model_list:
-      - model_name: gpt-4-azure
-        model_params:
-          model: gpt-4
-          context_limit: 128000
+      - model_name: gpt-4
+        model_id: gpt-4-deploy-name
+        context_limit: 128000
 ```
 
+**Key Tips:**
+- **Merging**: If `provider_name` matches a built-in provider, settings like `protocol` and `api_key` are inherited.
+- **Overriding**: Use the same `model_name` as a built-in model to override its parameters.
+- **Environment Variables**: Use `${VAR_NAME}` syntax for secrets.
+
 ##### Supported Protocols
 
 - `anthropic` - Anthropic Messages API

{klaude_code-2.3.0 → klaude_code-2.4.1}/README.md

@@ -133,95 +133,42 @@ Open in editor:
 klaude config
 ```
 
-##### Adding Models to Built-in Providers
+##### Model Configuration
 
-You can add custom models to existing built-in providers without redefining the entire provider. Just reference the `provider_name` and add your `model_list`:
+You can add custom models to built-in providers or define new ones. Configuration is inherited from built-in providers by matching `provider_name`.
 
 ```yaml
 # ~/.klaude/klaude-config.yaml
 provider_list:
-  - provider_name: openrouter  # Reference existing built-in provider
-    model_list:
-      - model_name: seed
-        model_params:
-          model: bytedance-seed/seed-1.6  # Model ID from OpenRouter
-          context_limit: 262000
-          cost:
-            input: 0.25
-            output: 2
-```
-
-**How merging works:**
-- Your models are merged with the built-in models for that provider
-- You only need `provider_name` and `model_list` - protocol, api_key, etc. are inherited from the built-in config
-- To override a built-in model, use the same `model_name` (e.g., `sonnet` to customize the built-in sonnet)
-
-**More examples:**
-
-```yaml
-provider_list:
-  # Add multiple models to OpenRouter
+  # Add/Override models for built-in OpenRouter provider
   - provider_name: openrouter
     model_list:
       - model_name: qwen-coder
-        model_params:
-          model: qwen/qwen-2.5-coder-32b-instruct
-          context_limit: 131072
-          cost:
-            input: 0.3
-            output: 0.9
-      - model_name: llama-405b
-        model_params:
-          model: meta-llama/llama-3.1-405b-instruct
-          context_limit: 131072
-          cost:
-            input: 0.8
-            output: 0.8
-
-  # Add models to Anthropic provider
-  - provider_name: anthropic
-    model_list:
-      - model_name: haiku@ant
-        model_params:
-          model: claude-3-5-haiku-20241022
-          context_limit: 200000
-          cost:
-            input: 1.0
-            output: 5.0
-```
-
-After adding models, run `klaude list` to verify they appear in the model list.
-
-##### Overriding Provider Settings
-
-Override provider-level settings (like api_key) while keeping built-in models:
-
-```yaml
-provider_list:
-  - provider_name: anthropic
-    api_key: sk-my-custom-key  # Override the default ${ANTHROPIC_API_KEY}
-    # Built-in models (sonnet, opus) are still available
-```
-
-##### Adding New Providers
-
-For providers not in the built-in list, you must specify `protocol`:
-
-```yaml
-provider_list:
-  - provider_name: my-azure-openai
+        model_id: qwen/qwen-2.5-coder-32b-instruct
+        context_limit: 131072
+        cost: { input: 0.3, output: 0.9 }
+      - model_name: sonnet # Override built-in sonnet params
+        model_id: anthropic/claude-3.5-sonnet
+        context_limit: 200000
+
+  # Add a completely new provider
+  - provider_name: my-azure
     protocol: openai
     api_key: ${AZURE_OPENAI_KEY}
     base_url: https://my-instance.openai.azure.com/
     is_azure: true
     azure_api_version: "2024-02-15-preview"
     model_list:
-      - model_name: gpt-4-azure
-        model_params:
-          model: gpt-4
-          context_limit: 128000
+      - model_name: gpt-4
+        model_id: gpt-4-deploy-name
+        context_limit: 128000
 ```
 
+**Key Tips:**
+- **Merging**: If `provider_name` matches a built-in provider, settings like `protocol` and `api_key` are inherited.
+- **Overriding**: Use the same `model_name` as a built-in model to override its parameters.
+- **Environment Variables**: Use `${VAR_NAME}` syntax for secrets.
+
 ##### Supported Protocols
 
 - `anthropic` - Anthropic Messages API

{klaude_code-2.3.0 → klaude_code-2.4.1}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "uv_build"
 
 [project]
 name = "klaude-code"
-version = "2.3.0"
+version = "2.4.1"
 description = "Minimal code agent CLI"
 readme = "README.md"
 requires-python = ">=3.13"

{klaude_code-2.3.0 → klaude_code-2.4.1}/src/klaude_code/cli/list_model.py

@@ -181,7 +181,7 @@ def format_env_var_display(value: str | None) -> Text:
 
 def _get_model_params_display(model: ModelConfig) -> list[Text]:
     """Get display elements for model parameters."""
-    param_strings = format_model_params(model.model_params)
+    param_strings = format_model_params(model)
     if param_strings:
         return [Text(s) for s in param_strings]
     return [Text("")]
@@ -286,7 +286,7 @@ def _build_models_table(
 
         if not provider_available:
             name = Text.assemble((prefix, ThemeKey.LINES), (model.model_name, "dim"))
-            model_id = Text(model.model_params.model or "", style="dim")
+            model_id = Text(model.model_id or "", style="dim")
             params = Text("(unavailable)", style="dim")
         else:
             # Build role tags for this model
@@ -305,7 +305,7 @@ def _build_models_table(
                 )
             else:
                 name = Text.assemble((prefix, ThemeKey.LINES), (model.model_name, ThemeKey.CONFIG_ITEM_NAME))
-            model_id = Text(model.model_params.model or "")
+            model_id = Text(model.model_id or "")
             params = Text(" · ").join(_get_model_params_display(model))
 
         models_table.add_row(name, model_id, params)

{klaude_code-2.3.0 → klaude_code-2.4.1}/src/klaude_code/cli/main.py

@@ -124,7 +124,7 @@ def main_callback(
             raise typer.Exit(2)
 
         from klaude_code.app.runtime import AppInitConfig
-        from klaude_code.tui.command.model_select import ModelSelectStatus, select_model_interactive
+        from klaude_code.tui.command.model_picker import ModelSelectStatus, select_model_interactive
         from klaude_code.tui.runner import run_interactive
 
         update_terminal_title()
@@ -193,7 +193,7 @@ def main_callback(
                     matches = [
                         m.selector
                         for m in cfg.iter_model_entries()
-                        if (m.model_params.model or "").strip().lower() == raw_model.lower()
+                        if (m.model_id or "").strip().lower() == raw_model.lower()
                     ]
                     if len(matches) == 1:
                         chosen_model = matches[0]

klaude_code-2.4.1/src/klaude_code/config/assets/builtin_config.yaml

@@ -0,0 +1,236 @@
+---
+# Built-in provider and model configurations
+# Users can start using klaude by simply setting environment variables
+# (ANTHROPIC_API_KEY, OPENAI_API_KEY, etc.) without manual configuration.
+provider_list:
+- provider_name: anthropic
+  protocol: anthropic
+  api_key: ${ANTHROPIC_API_KEY}
+  model_list:
+  - model_name: sonnet
+    model_id: claude-sonnet-4-5-20250929
+    context_limit: 200000
+    provider_routing:
+      sort: throughput
+    cost: {input: 3, output: 15, cache_read: 0.3, cache_write: 3.75}
+  - model_name: opus
+    model_id: claude-opus-4-5-20251101
+    context_limit: 200000
+    verbosity: high
+    thinking:
+      type: enabled
+      budget_tokens: 2048
+    cost: {input: 5, output: 25, cache_read: 0.5, cache_write: 6.25}
+- provider_name: openai
+  protocol: responses
+  api_key: ${OPENAI_API_KEY}
+  model_list:
+  - model_name: gpt-5.2
+    model_id: gpt-5.2
+    max_tokens: 128000
+    context_limit: 400000
+    verbosity: high
+    thinking:
+      reasoning_effort: high
+    cost: {input: 1.75, output: 14, cache_read: 0.17}
+- provider_name: openrouter
+  protocol: openrouter
+  api_key: ${OPENROUTER_API_KEY}
+  model_list:
+  - model_name: gpt-5.1-codex-max
+    model_id: openai/gpt-5.1-codex-max
+    max_tokens: 128000
+    context_limit: 400000
+    thinking:
+      reasoning_effort: medium
+    cost: {input: 1.25, output: 10, cache_read: 0.13}
+  - model_name: gpt-5.2
+    model_id: openai/gpt-5.2
+    max_tokens: 128000
+    context_limit: 400000
+    verbosity: high
+    thinking:
+      reasoning_effort: high
+    cost: {input: 1.75, output: 14, cache_read: 0.17}
+  - model_name: gpt-5.2-medium
+    model_id: openai/gpt-5.2
+    max_tokens: 128000
+    context_limit: 400000
+    verbosity: high
+    thinking:
+      reasoning_effort: medium
+    cost: {input: 1.75, output: 14, cache_read: 0.17}
+  - model_name: gpt-5.2-low
+    model_id: openai/gpt-5.2
+    max_tokens: 128000
+    context_limit: 400000
+    verbosity: low
+    thinking:
+      reasoning_effort: low
+    cost: {input: 1.75, output: 14, cache_read: 0.17}
+  - model_name: gpt-5.2-fast
+    model_id: openai/gpt-5.2
+    max_tokens: 128000
+    context_limit: 400000
+    verbosity: low
+    thinking:
+      reasoning_effort: none
+    cost: {input: 1.75, output: 14, cache_read: 0.17}
+  - model_name: kimi
+    model_id: moonshotai/kimi-k2-thinking
+    context_limit: 262144
+    provider_routing:
+      only:
+      - moonshotai/turbo
+    cost: {input: 0.6, output: 2.5, cache_read: 0.15}
+  - model_name: haiku
+    model_id: anthropic/claude-haiku-4.5
+    context_limit: 200000
+    cost: {input: 1, output: 5, cache_read: 0.1, cache_write: 1.25}
+  - model_name: sonnet
+    model_id: anthropic/claude-4.5-sonnet
+    context_limit: 200000
+    provider_routing:
+      sort: throughput
+    cost: {input: 3, output: 15, cache_read: 0.3, cache_write: 3.75}
+  - model_name: opus
+    model_id: anthropic/claude-4.5-opus
+    context_limit: 200000
+    verbosity: high
+    thinking:
+      type: enabled
+      budget_tokens: 2048
+    cost: {input: 5, output: 25, cache_read: 0.5, cache_write: 6.25}
+  - model_name: gemini-pro
+    model_id: google/gemini-3-pro-preview
+    context_limit: 1048576
+    thinking:
+      reasoning_effort: high
+    cost: {input: 2, output: 12, cache_read: 0.2}
+  - model_name: gemini-flash
+    model_id: google/gemini-3-flash-preview
+    context_limit: 1048576
+    thinking:
+      reasoning_effort: medium
+    cost: {input: 0.5, output: 3, cache_read: 0.05}
+  - model_name: nano-banana-pro
+    model_id: google/gemini-3-pro-image-preview
+    context_limit: 66000
+    modalities:
+    - image
+    - text
+    cost: {input: 2, output: 12, cache_read: 0.2, image: 120}
+  - model_name: nano-banana
+    model_id: google/gemini-2.5-flash-image
+    context_limit: 33000
+    modalities:
+    - image
+    - text
+    cost: {input: 0.3, output: 2.5, cache_read: 0.03, image: 30}
+  - model_name: grok
+    model_id: x-ai/grok-4.1-fast
+    context_limit: 2000000
+    thinking:
+      type: enabled
+      budget_tokens: 2048
+    cost: {input: 0.2, output: 0.5, cache_read: 0.05}
+  - model_name: minimax
+    model_id: minimax/minimax-m2.1
+    context_limit: 204800
+    cost: {input: 0.3, output: 1.2, cache_read: 0.03}
+  - model_name: glm
+    model_id: z-ai/glm-4.7
+    context_limit: 200000
+    provider_routing:
+      only:
+      - z-ai
+    cost: {input: 0.44, output: 1.74, cache_read: 0.04}
+  - model_name: seedream
+    model_id: bytedance-seed/seedream-4.5
+    context_limit: 4000
+    cost: {input: 0, output: 9.581, image: 9.581}
+    modalities:
+    - image
+    - text
+- provider_name: google
+  protocol: google
+  api_key: ${GOOGLE_API_KEY}
+  model_list:
+  - model_name: gemini-pro
+    model_id: gemini-3-pro-preview
+    context_limit: 1048576
+    cost: {input: 2, output: 12, cache_read: 0.2}
+  - model_name: gemini-flash
+    model_id: gemini-3-flash-preview
+    context_limit: 1048576
+    cost: {input: 0.5, output: 3, cache_read: 0.05}
+  - model_name: nano-banana-pro
+    model_id: gemini-3-pro-image-preview
+    context_limit: 66000
+    modalities:
+    - image
+    - text
+    cost: {input: 2, output: 12, cache_read: 0.2, image: 120}
+- provider_name: bedrock
+  protocol: bedrock
+  aws_access_key: ${AWS_ACCESS_KEY_ID}
+  aws_secret_key: ${AWS_SECRET_ACCESS_KEY}
+  aws_region: ${AWS_REGION}
+  model_list:
+  - model_name: sonnet
+    model_id: us.anthropic.claude-sonnet-4-5-20250929-v1:0
+    context_limit: 200000
+    cost: {input: 3, output: 15, cache_read: 0.3, cache_write: 3.75}
+- provider_name: deepseek
+  protocol: anthropic
+  api_key: ${DEEPSEEK_API_KEY}
+  base_url: https://api.deepseek.com/anthropic
+  model_list:
+  - model_name: deepseek
+    model_id: deepseek-reasoner
+    context_limit: 128000
+    thinking:
+      type: enabled
+      budget_tokens: 2048
+    cost: {input: 2, output: 3, cache_read: 0.2, currency: CNY}
+- provider_name: moonshot
+  protocol: anthropic
+  api_key: ${MOONSHOT_API_KEY}
+  base_url: https://api.moonshot.cn/anthropic
+  model_list:
+  - model_name: kimi
+    model_id: kimi-k2-thinking
+    context_limit: 262144
+    thinking:
+      type: enabled
+      budget_tokens: 8192
+    cost: {input: 4, output: 16, cache_read: 1, currency: CNY}
+- provider_name: claude-max
+  protocol: claude_oauth
+  model_list:
+  - model_name: sonnet
+    model_id: claude-sonnet-4-5-20250929
+    context_limit: 200000
+    cost: {input: 3, output: 15, cache_read: 0.3, cache_write: 3.75}
+  - model_name: opus
+    model_id: claude-opus-4-5-20251101
+    context_limit: 200000
+    verbosity: high
+    thinking:
+      type: enabled
+      budget_tokens: 2048
+    cost: {input: 5, output: 25, cache_read: 0.5, cache_write: 6.25}
+  - model_name: haiku
+    model_id: claude-haiku-4-5-20251001
+    context_limit: 200000
+    cost: {input: 1, output: 5, cache_read: 0.1, cache_write: 1.25}
+- provider_name: codex
+  protocol: codex_oauth
+  model_list:
+  - model_name: gpt-5.2-codex
+    model_id: gpt-5.2-codex
+    thinking:
+      reasoning_effort: medium
+    context_limit: 400000
+    max_tokens: 128000
+    cost: {input: 1.75, output: 14, cache_read: 0.17}

{klaude_code-2.3.0 → klaude_code-2.4.1}/src/klaude_code/config/config.py

@@ -58,9 +58,10 @@ config_path = Path.home() / ".klaude" / "klaude-config.yaml"
 example_config_path = Path.home() / ".klaude" / "klaude-config.example.yaml"
 
 
-class ModelConfig(BaseModel):
+class ModelConfig(llm_param.LLMConfigModelParameter):
+    """Model configuration that flattens LLMConfigModelParameter fields."""
+
     model_name: str
-    model_params: llm_param.LLMConfigModelParameter
 
 
 class ProviderConfig(llm_param.LLMConfigProviderParameter):
@@ -135,10 +136,11 @@ class UserProviderConfig(BaseModel):
     model_list: list[ModelConfig] = Field(default_factory=lambda: [])
 
 
-class ModelEntry(BaseModel):
+class ModelEntry(llm_param.LLMConfigModelParameter):
+    """Model entry with provider info, flattens LLMConfigModelParameter fields."""
+
     model_name: str
     provider: str
-    model_params: llm_param.LLMConfigModelParameter
 
     @property
     def selector(self) -> str:
@@ -325,7 +327,7 @@ class Config(BaseModel):
                 provider_dump["api_key"] = api_key
                 return llm_param.LLMConfigParameter(
                     **provider_dump,
-                    **model.model_params.model_dump(),
+                    **model.model_dump(exclude={"model_name"}),
                 )
 
         raise ValueError(f"Unknown model: {model_name}")
@@ -340,7 +342,7 @@ class Config(BaseModel):
             ModelEntry(
                 model_name=model.model_name,
                 provider=provider.provider_name,
-                model_params=model.model_params,
+                **model.model_dump(exclude={"model_name"}),
             )
             for provider in self.provider_list
             if not only_available or not provider.is_api_key_missing()
@@ -350,7 +352,7 @@ class Config(BaseModel):
     def has_available_image_model(self) -> bool:
         """Check if any image generation model is available."""
         for entry in self.iter_model_entries(only_available=True):
-            if entry.model_params.modalities and "image" in entry.model_params.modalities:
+            if entry.modalities and "image" in entry.modalities:
                 return True
         return False
 
@@ -364,7 +366,7 @@ class Config(BaseModel):
     def get_first_available_image_model(self) -> str | None:
         """Get the first available image generation model, or None."""
         for entry in self.iter_model_entries(only_available=True):
-            if entry.model_params.modalities and "image" in entry.model_params.modalities:
+            if entry.modalities and "image" in entry.modalities:
                 return entry.model_name
         return None
 
@@ -409,15 +411,13 @@ def get_example_config() -> UserConfig:
                 model_list=[
                     ModelConfig(
                         model_name="my-model",
-                        model_params=llm_param.LLMConfigModelParameter(
-                            model="model-id-from-provider",
-                            max_tokens=16000,
-                            context_limit=200000,
-                            cost=llm_param.Cost(
-                                input=1,
-                                output=10,
-                                cache_read=0.1,
-                            ),
+                        model_id="model-id-from-provider",
+                        max_tokens=16000,
+                        context_limit=200000,
+                        cost=llm_param.Cost(
+                            input=1,
+                            output=10,
+                            cache_read=0.1,
                         ),
                     ),
                 ],

klaude_code-2.3.0/src/klaude_code/config/select_model.py → klaude_code-2.4.1/src/klaude_code/config/model_matcher.py

@@ -51,7 +51,7 @@ def match_model_from_config(preferred: str | None = None) -> ModelMatchResult:
     # Only show models from providers with valid API keys
     models: list[ModelEntry] = sorted(
         config.iter_model_entries(only_available=True),
-        key=lambda m: (m.model_name.lower(), m.provider.lower()),
+        key=lambda m: (m.provider.lower(), m.model_name.lower()),
     )
 
     if not models:
@@ -86,13 +86,13 @@ def match_model_from_config(preferred: str | None = None) -> ModelMatchResult:
             return ModelMatchResult(matched_model=None, filtered_models=exact_base_matches, filter_hint=filter_hint)
 
         preferred_lower = preferred.lower()
-        # Case-insensitive exact match (selector/model_name/model_params.model)
+        # Case-insensitive exact match (selector/model_name/model_id)
         exact_ci_matches = [
             m
             for m in models
             if preferred_lower == m.selector.lower()
             or preferred_lower == m.model_name.lower()
-            or preferred_lower == (m.model_params.model or "").lower()
+            or preferred_lower == (m.model_id or "").lower()
         ]
         if len(exact_ci_matches) == 1:
             return ModelMatchResult(
@@ -110,7 +110,7 @@ def match_model_from_config(preferred: str | None = None) -> ModelMatchResult:
                 for m in models
                 if preferred_norm == _normalize_model_key(m.selector)
                 or preferred_norm == _normalize_model_key(m.model_name)
-                or preferred_norm == _normalize_model_key(m.model_params.model or "")
+                or preferred_norm == _normalize_model_key(m.model_id or "")
             ]
             if len(normalized_matches) == 1:
                 return ModelMatchResult(
@@ -125,7 +125,7 @@ def match_model_from_config(preferred: str | None = None) -> ModelMatchResult:
                     for m in models
                     if preferred_norm in _normalize_model_key(m.selector)
                     or preferred_norm in _normalize_model_key(m.model_name)
-                    or preferred_norm in _normalize_model_key(m.model_params.model or "")
+                    or preferred_norm in _normalize_model_key(m.model_id or "")
                 ]
                 if len(normalized_matches) == 1:
                     return ModelMatchResult(
@@ -134,14 +134,14 @@ def match_model_from_config(preferred: str | None = None) -> ModelMatchResult:
                         filter_hint=None,
                     )
 
-        # Partial match (case-insensitive) on model_name or model_params.model.
+        # Partial match (case-insensitive) on model_name or model_id.
         # If normalized matching found candidates (even if multiple), prefer those as the filter set.
         matches = normalized_matches or [
             m
             for m in models
             if preferred_lower in m.selector.lower()
             or preferred_lower in m.model_name.lower()
-            or preferred_lower in (m.model_params.model or "").lower()
+            or preferred_lower in (m.model_id or "").lower()
         ]
         if len(matches) == 1:
             return ModelMatchResult(matched_model=matches[0].selector, filtered_models=models, filter_hint=None)