superqode 0.1.5__py3-none-any.whl

superqode/providers/models.py

@@ -0,0 +1,1016 @@
+"""
+SuperQode Model Database - Model pricing, features, and metadata.
+
+Provides detailed information about LLM models including:
+- Pricing (input/output per 1M tokens)
+- Context window size
+- Feature support (tools, vision, etc.)
+- Recommendations
+
+Usage:
+    from superqode.providers.models import get_model_info, MODELS
+
+    info = get_model_info("anthropic", "claude-sonnet-4")
+    print(f"Price: ${info.input_price}/${info.output_price} per 1M tokens")
+"""
+
+from __future__ import annotations
+
+from dataclasses import dataclass, field
+from enum import Enum, auto
+from typing import Dict, List, Optional, Tuple
+
+
+# ============================================================================
+# MODEL INFO
+# ============================================================================
+
+
+class ModelCapability(Enum):
+    """Model capabilities."""
+
+    TOOLS = auto()  # Function calling / tools
+    VISION = auto()  # Image input
+    STREAMING = auto()  # Streaming output
+    JSON_MODE = auto()  # Structured JSON output
+    REASONING = auto()  # Extended thinking / reasoning
+    CODE = auto()  # Optimized for code
+    LONG_CONTEXT = auto()  # > 100K context
+
+
+@dataclass
+class ModelInfo:
+    """Detailed model information."""
+
+    id: str  # Model identifier
+    name: str  # Human-readable name
+    provider: str  # Provider ID
+
+    # Pricing (per 1M tokens, USD)
+    input_price: float = 0.0
+    output_price: float = 0.0
+
+    # Context
+    context_window: int = 128000  # Max tokens
+    max_output: int = 4096  # Max output tokens
+
+    # Capabilities
+    capabilities: List[ModelCapability] = field(default_factory=list)
+
+    # Metadata
+    description: str = ""
+    recommended_for: List[str] = field(default_factory=list)
+    released: str = ""  # Release date
+
+    @property
+    def supports_tools(self) -> bool:
+        return ModelCapability.TOOLS in self.capabilities
+
+    @property
+    def supports_vision(self) -> bool:
+        return ModelCapability.VISION in self.capabilities
+
+    @property
+    def supports_reasoning(self) -> bool:
+        return ModelCapability.REASONING in self.capabilities
+
+    @property
+    def is_code_optimized(self) -> bool:
+        return ModelCapability.CODE in self.capabilities
+
+    @property
+    def price_display(self) -> str:
+        """Display-friendly pricing."""
+        if self.input_price == 0 and self.output_price == 0:
+            return "Free"
+        return f"${self.input_price:.2f}/${self.output_price:.2f}"
+
+    @property
+    def context_display(self) -> str:
+        """Display-friendly context window."""
+        if self.context_window >= 1000000:
+            return f"{self.context_window // 1000000}M"
+        elif self.context_window >= 1000:
+            return f"{self.context_window // 1000}K"
+        return str(self.context_window)
+
+    def estimate_cost(self, input_tokens: int, output_tokens: int) -> float:
+        """Estimate cost for given token counts."""
+        input_cost = (input_tokens / 1_000_000) * self.input_price
+        output_cost = (output_tokens / 1_000_000) * self.output_price
+        return input_cost + output_cost
+
+
+# ============================================================================
+# MODEL DATABASE
+# ============================================================================
+
+MODELS: Dict[str, Dict[str, ModelInfo]] = {
+    # =========================================================================
+    # ANTHROPIC
+    # =========================================================================
+    "anthropic": {
+        "claude-opus-4-5-20251101": ModelInfo(
+            id="claude-opus-4-5-20251101",
+            name="Claude Opus 4.5",
+            provider="anthropic",
+            input_price=15.0,
+            output_price=75.0,
+            context_window=200000,
+            max_output=8192,
+            capabilities=[
+                ModelCapability.TOOLS,
+                ModelCapability.VISION,
+                ModelCapability.STREAMING,
+                ModelCapability.JSON_MODE,
+                ModelCapability.REASONING,
+                ModelCapability.CODE,
+                ModelCapability.LONG_CONTEXT,
+            ],
+            description="Most capable Claude model - latest flagship",
+            recommended_for=["complex reasoning", "research", "difficult coding"],
+            released="2025-11",
+        ),
+        "claude-sonnet-4-5-20250929": ModelInfo(
+            id="claude-sonnet-4-5-20250929",
+            name="Claude Sonnet 4.5",
+            provider="anthropic",
+            input_price=3.0,
+            output_price=15.0,
+            context_window=200000,
+            max_output=8192,
+            capabilities=[
+                ModelCapability.TOOLS,
+                ModelCapability.VISION,
+                ModelCapability.STREAMING,
+                ModelCapability.JSON_MODE,
+                ModelCapability.CODE,
+                ModelCapability.LONG_CONTEXT,
+            ],
+            description="Best balance of intelligence and speed",
+            recommended_for=["coding", "analysis", "general"],
+            released="2025-09",
+        ),
+        "claude-haiku-4-5-20251001": ModelInfo(
+            id="claude-haiku-4-5-20251001",
+            name="Claude Haiku 4.5",
+            provider="anthropic",
+            input_price=0.25,
+            output_price=1.25,
+            context_window=200000,
+            max_output=8192,
+            capabilities=[
+                ModelCapability.TOOLS,
+                ModelCapability.VISION,
+                ModelCapability.STREAMING,
+                ModelCapability.JSON_MODE,
+                ModelCapability.LONG_CONTEXT,
+            ],
+            description="Fastest and most cost-effective",
+            recommended_for=["quick tasks", "high volume"],
+            released="2025-10",
+        ),
+        "claude-sonnet-4-20250514": ModelInfo(
+            id="claude-sonnet-4-20250514",
+            name="Claude Sonnet 4",
+            provider="anthropic",
+            input_price=3.0,
+            output_price=15.0,
+            context_window=200000,
+            max_output=8192,
+            capabilities=[
+                ModelCapability.TOOLS,
+                ModelCapability.VISION,
+                ModelCapability.STREAMING,
+                ModelCapability.JSON_MODE,
+                ModelCapability.CODE,
+                ModelCapability.LONG_CONTEXT,
+            ],
+            description="Previous Sonnet generation",
+            recommended_for=["coding", "analysis", "general"],
+            released="2025-05",
+        ),
+        "claude-opus-4-20250514": ModelInfo(
+            id="claude-opus-4-20250514",
+            name="Claude Opus 4",
+            provider="anthropic",
+            input_price=15.0,
+            output_price=75.0,
+            context_window=200000,
+            max_output=8192,
+            capabilities=[
+                ModelCapability.TOOLS,
+                ModelCapability.VISION,
+                ModelCapability.STREAMING,
+                ModelCapability.JSON_MODE,
+                ModelCapability.REASONING,
+                ModelCapability.CODE,
+                ModelCapability.LONG_CONTEXT,
+            ],
+            description="Previous Opus generation",
+            recommended_for=["complex reasoning", "research", "difficult coding"],
+            released="2025-05",
+        ),
+        "claude-haiku-4-20250514": ModelInfo(
+            id="claude-haiku-4-20250514",
+            name="Claude Haiku 4",
+            provider="anthropic",
+            input_price=0.25,
+            output_price=1.25,
+            context_window=200000,
+            max_output=8192,
+            capabilities=[
+                ModelCapability.TOOLS,
+                ModelCapability.VISION,
+                ModelCapability.STREAMING,
+                ModelCapability.JSON_MODE,
+                ModelCapability.LONG_CONTEXT,
+            ],
+            description="Previous Haiku generation",
+            recommended_for=["quick tasks", "high volume"],
+            released="2025-05",
+        ),
+    },
+    # =========================================================================
+    # OPENAI
+    # =========================================================================
+    "openai": {
+        "gpt-5.2": ModelInfo(
+            id="gpt-5.2",
+            name="GPT-5.2",
+            provider="openai",
+            input_price=5.0,
+            output_price=20.0,
+            context_window=256000,
+            max_output=32768,
+            capabilities=[
+                ModelCapability.TOOLS,
+                ModelCapability.VISION,
+                ModelCapability.STREAMING,
+                ModelCapability.JSON_MODE,
+                ModelCapability.REASONING,
+                ModelCapability.CODE,
+                ModelCapability.LONG_CONTEXT,
+            ],
+            description="Latest GPT-5 flagship with reasoning",
+            recommended_for=["complex reasoning", "coding", "research"],
+            released="2025-12",
+        ),
+        "gpt-5.2-pro": ModelInfo(
+            id="gpt-5.2-pro",
+            name="GPT-5.2 Pro",
+            provider="openai",
+            input_price=6.0,
+            output_price=24.0,
+            context_window=256000,
+            max_output=32768,
+            capabilities=[
+                ModelCapability.TOOLS,
+                ModelCapability.VISION,
+                ModelCapability.STREAMING,
+                ModelCapability.JSON_MODE,
+                ModelCapability.REASONING,
+                ModelCapability.CODE,
+                ModelCapability.LONG_CONTEXT,
+            ],
+            description="GPT-5.2 Pro variant - enhanced capabilities",
+            recommended_for=["complex reasoning", "coding", "research"],
+            released="2025-12",
+        ),
+        "gpt-5.2-codex": ModelInfo(
+            id="gpt-5.2-codex",
+            name="GPT-5.2 Codex",
+            provider="openai",
+            input_price=5.5,
+            output_price=22.0,
+            context_window=256000,
+            max_output=32768,
+            capabilities=[
+                ModelCapability.TOOLS,
+                ModelCapability.VISION,
+                ModelCapability.STREAMING,
+                ModelCapability.JSON_MODE,
+                ModelCapability.REASONING,
+                ModelCapability.CODE,
+                ModelCapability.LONG_CONTEXT,
+            ],
+            description="GPT-5.2 Codex variant - optimized for code",
+            recommended_for=["coding", "code generation", "code review"],
+            released="2025-12",
+        ),
+        "gpt-5.1": ModelInfo(
+            id="gpt-5.1",
+            name="GPT-5.1",
+            provider="openai",
+            input_price=4.0,
+            output_price=16.0,
+            context_window=200000,
+            max_output=32768,
+            capabilities=[
+                ModelCapability.TOOLS,
+                ModelCapability.VISION,
+                ModelCapability.STREAMING,
+                ModelCapability.JSON_MODE,
+                ModelCapability.REASONING,
+                ModelCapability.CODE,
+                ModelCapability.LONG_CONTEXT,
+            ],
+            description="GPT-5 series - highly capable",
+            recommended_for=["general", "coding", "analysis"],
+            released="2025-11",
+        ),
+        "gpt-5.1-codex": ModelInfo(
+            id="gpt-5.1-codex",
+            name="GPT-5.1 Codex",
+            provider="openai",
+            input_price=4.5,
+            output_price=18.0,
+            context_window=200000,
+            max_output=32768,
+            capabilities=[
+                ModelCapability.TOOLS,
+                ModelCapability.VISION,
+                ModelCapability.STREAMING,
+                ModelCapability.JSON_MODE,
+                ModelCapability.REASONING,
+                ModelCapability.CODE,
+                ModelCapability.LONG_CONTEXT,
+            ],
+            description="GPT-5.1 Codex variant - optimized for code",
+            recommended_for=["coding", "code generation"],
+            released="2025-11",
+        ),
+        "gpt-5.1-codex-mini": ModelInfo(
+            id="gpt-5.1-codex-mini",
+            name="GPT-5.1 Codex Mini",
+            provider="openai",
+            input_price=2.0,
+            output_price=8.0,
+            context_window=200000,
+            max_output=16384,
+            capabilities=[
+                ModelCapability.TOOLS,
+                ModelCapability.VISION,
+                ModelCapability.STREAMING,
+                ModelCapability.JSON_MODE,
+                ModelCapability.CODE,
+                ModelCapability.LONG_CONTEXT,
+            ],
+            description="GPT-5.1 Codex Mini - fast and efficient for code",
+            recommended_for=["quick coding", "code completion"],
+            released="2025-11",
+        ),
+        "gpt-4o-2024-11-20": ModelInfo(
+            id="gpt-4o-2024-11-20",
+            name="GPT-4o (Nov 2024)",
+            provider="openai",
+            input_price=2.50,
+            output_price=10.0,
+            context_window=128000,
+            max_output=16384,
+            capabilities=[
+                ModelCapability.TOOLS,
+                ModelCapability.VISION,
+                ModelCapability.STREAMING,
+                ModelCapability.JSON_MODE,
+                ModelCapability.CODE,
+                ModelCapability.LONG_CONTEXT,
+            ],
+            description="GPT-4o latest stable version",
+            recommended_for=["general", "coding", "vision"],
+            released="2024-11",
+        ),
+        "gpt-4o": ModelInfo(
+            id="gpt-4o",
+            name="GPT-4o",
+            provider="openai",
+            input_price=2.50,
+            output_price=10.0,
+            context_window=128000,
+            max_output=16384,
+            capabilities=[
+                ModelCapability.TOOLS,
+                ModelCapability.VISION,
+                ModelCapability.STREAMING,
+                ModelCapability.JSON_MODE,
+                ModelCapability.CODE,
+                ModelCapability.LONG_CONTEXT,
+            ],
+            description="Most capable GPT-4 variant",
+            recommended_for=["general", "coding", "vision"],
+            released="2024-05",
+        ),
+        "gpt-4o-mini": ModelInfo(
+            id="gpt-4o-mini",
+            name="GPT-4o Mini",
+            provider="openai",
+            input_price=0.15,
+            output_price=0.60,
+            context_window=128000,
+            max_output=16384,
+            capabilities=[
+                ModelCapability.TOOLS,
+                ModelCapability.VISION,
+                ModelCapability.STREAMING,
+                ModelCapability.JSON_MODE,
+                ModelCapability.LONG_CONTEXT,
+            ],
+            description="Fast and cost-effective",
+            recommended_for=["quick tasks", "high volume"],
+            released="2024-07",
+        ),
+        "o1": ModelInfo(
+            id="o1",
+            name="o1",
+            provider="openai",
+            input_price=15.0,
+            output_price=60.0,
+            context_window=200000,
+            max_output=100000,
+            capabilities=[
+                ModelCapability.REASONING,
+                ModelCapability.CODE,
+                ModelCapability.LONG_CONTEXT,
+            ],
+            description="Advanced reasoning model",
+            recommended_for=["complex reasoning", "math", "science"],
+            released="2024-09",
+        ),
+        "o1-mini": ModelInfo(
+            id="o1-mini",
+            name="o1-mini",
+            provider="openai",
+            input_price=3.0,
+            output_price=12.0,
+            context_window=128000,
+            max_output=65536,
+            capabilities=[
+                ModelCapability.REASONING,
+                ModelCapability.CODE,
+                ModelCapability.LONG_CONTEXT,
+            ],
+            description="Smaller reasoning model",
+            recommended_for=["coding", "math"],
+            released="2024-09",
+        ),
+    },
+    # =========================================================================
+    # GOOGLE
+    # =========================================================================
+    "google": {
+        "gemini-3-pro-preview": ModelInfo(
+            id="gemini-3-pro-preview",
+            name="Gemini 3 Pro Preview",
+            provider="google",
+            input_price=2.0,
+            output_price=8.0,
+            context_window=2000000,
+            max_output=16384,
+            capabilities=[
+                ModelCapability.TOOLS,
+                ModelCapability.VISION,
+                ModelCapability.STREAMING,
+                ModelCapability.JSON_MODE,
+                ModelCapability.REASONING,
+                ModelCapability.CODE,
+                ModelCapability.LONG_CONTEXT,
+            ],
+            description="Latest Gemini 3 flagship - most capable (2M context)",
+            recommended_for=["complex reasoning", "large codebases", "research"],
+            released="2025-12",
+        ),
+        "gemini-3-flash-preview": ModelInfo(
+            id="gemini-3-flash-preview",
+            name="Gemini 3 Flash Preview",
+            provider="google",
+            input_price=0.15,
+            output_price=0.60,
+            context_window=1000000,
+            max_output=8192,
+            capabilities=[
+                ModelCapability.TOOLS,
+                ModelCapability.VISION,
+                ModelCapability.STREAMING,
+                ModelCapability.JSON_MODE,
+                ModelCapability.CODE,
+                ModelCapability.LONG_CONTEXT,
+            ],
+            description="Gemini 3 fast model - 1M context (Latest)",
+            recommended_for=["quick tasks", "high volume", "coding"],
+            released="2025-12",
+        ),
+        "gemini-2.5-pro": ModelInfo(
+            id="gemini-2.5-pro",
+            name="Gemini 2.5 Pro",
+            provider="google",
+            input_price=1.25,
+            output_price=5.0,
+            context_window=2000000,
+            max_output=8192,
+            capabilities=[
+                ModelCapability.TOOLS,
+                ModelCapability.VISION,
+                ModelCapability.STREAMING,
+                ModelCapability.JSON_MODE,
+                ModelCapability.REASONING,
+                ModelCapability.CODE,
+                ModelCapability.LONG_CONTEXT,
+            ],
+            description="Gemini 2.5 Pro with 2M context",
+            recommended_for=["large codebases", "long documents"],
+            released="2025-01",
+        ),
+        "gemini-2.5-flash": ModelInfo(
+            id="gemini-2.5-flash",
+            name="Gemini 2.5 Flash",
+            provider="google",
+            input_price=0.075,
+            output_price=0.30,
+            context_window=1000000,
+            max_output=8192,
+            capabilities=[
+                ModelCapability.TOOLS,
+                ModelCapability.VISION,
+                ModelCapability.STREAMING,
+                ModelCapability.JSON_MODE,
+                ModelCapability.CODE,
+                ModelCapability.LONG_CONTEXT,
+            ],
+            description="Fast and efficient with 1M context",
+            recommended_for=["quick tasks", "high volume"],
+            released="2025-01",
+        ),
+        "gemini-2.0-flash": ModelInfo(
+            id="gemini-2.0-flash",
+            name="Gemini 2.0 Flash",
+            provider="google",
+            input_price=0.10,
+            output_price=0.40,
+            context_window=1000000,
+            max_output=8192,
+            capabilities=[
+                ModelCapability.TOOLS,
+                ModelCapability.VISION,
+                ModelCapability.STREAMING,
+                ModelCapability.JSON_MODE,
+                ModelCapability.CODE,
+                ModelCapability.LONG_CONTEXT,
+            ],
+            description="Previous Flash generation",
+            recommended_for=["general tasks"],
+            released="2024-12",
+        ),
+    },
+    # =========================================================================
+    # DEEPSEEK
+    # =========================================================================
+    "deepseek": {
+        "deepseek-ai/DeepSeek-V3.2": ModelInfo(
+            id="deepseek-ai/DeepSeek-V3.2",
+            name="DeepSeek V3.2",
+            provider="deepseek",
+            input_price=0.27,
+            output_price=1.10,
+            context_window=128000,
+            max_output=16384,
+            capabilities=[
+                ModelCapability.TOOLS,
+                ModelCapability.STREAMING,
+                ModelCapability.JSON_MODE,
+                ModelCapability.REASONING,
+                ModelCapability.CODE,
+                ModelCapability.LONG_CONTEXT,
+            ],
+            description="Latest DeepSeek V3.2 - most capable",
+            recommended_for=["complex reasoning", "coding", "research"],
+            released="2025-12",
+        ),
+        "deepseek-ai/DeepSeek-R1": ModelInfo(
+            id="deepseek-ai/DeepSeek-R1",
+            name="DeepSeek R1",
+            provider="deepseek",
+            input_price=0.55,
+            output_price=2.19,
+            context_window=64000,
+            max_output=8192,
+            capabilities=[
+                ModelCapability.STREAMING,
+                ModelCapability.REASONING,
+                ModelCapability.CODE,
+            ],
+            description="Advanced reasoning model - R1 series",
+            recommended_for=["complex reasoning", "math", "coding"],
+            released="2025-01",
+        ),
+        "deepseek-chat": ModelInfo(
+            id="deepseek-chat",
+            name="DeepSeek Chat (V3)",
+            provider="deepseek",
+            input_price=0.14,
+            output_price=0.28,
+            context_window=64000,
+            max_output=8192,
+            capabilities=[
+                ModelCapability.TOOLS,
+                ModelCapability.STREAMING,
+                ModelCapability.JSON_MODE,
+                ModelCapability.CODE,
+            ],
+            description="Very cost-effective general model",
+            recommended_for=["general", "budget-conscious"],
+            released="2024-12",
+        ),
+        "deepseek-coder": ModelInfo(
+            id="deepseek-coder",
+            name="DeepSeek Coder",
+            provider="deepseek",
+            input_price=0.14,
+            output_price=0.28,
+            context_window=64000,
+            max_output=8192,
+            capabilities=[
+                ModelCapability.TOOLS,
+                ModelCapability.STREAMING,
+                ModelCapability.JSON_MODE,
+                ModelCapability.CODE,
+            ],
+            description="Specialized for coding tasks",
+            recommended_for=["coding", "code review"],
+            released="2024-12",
+        ),
+        "deepseek-reasoner": ModelInfo(
+            id="deepseek-reasoner",
+            name="DeepSeek Reasoner",
+            provider="deepseek",
+            input_price=0.55,
+            output_price=2.19,
+            context_window=64000,
+            max_output=8192,
+            capabilities=[
+                ModelCapability.STREAMING,
+                ModelCapability.REASONING,
+                ModelCapability.CODE,
+            ],
+            description="Advanced reasoning model",
+            recommended_for=["complex reasoning", "math"],
+            released="2025-01",
+        ),
+    },
+    # =========================================================================
+    # GROQ
+    # =========================================================================
+    "groq": {
+        "llama-3.3-70b-versatile": ModelInfo(
+            id="llama-3.3-70b-versatile",
+            name="Llama 3.3 70B",
+            provider="groq",
+            input_price=0.59,
+            output_price=0.79,
+            context_window=128000,
+            max_output=32768,
+            capabilities=[
+                ModelCapability.TOOLS,
+                ModelCapability.STREAMING,
+                ModelCapability.JSON_MODE,
+                ModelCapability.CODE,
+                ModelCapability.LONG_CONTEXT,
+            ],
+            description="Very fast inference via Groq LPU",
+            recommended_for=["speed-critical", "coding"],
+            released="2024-12",
+        ),
+        "llama-3.1-8b-instant": ModelInfo(
+            id="llama-3.1-8b-instant",
+            name="Llama 3.1 8B Instant",
+            provider="groq",
+            input_price=0.05,
+            output_price=0.08,
+            context_window=128000,
+            max_output=8192,
+            capabilities=[
+                ModelCapability.TOOLS,
+                ModelCapability.STREAMING,
+                ModelCapability.LONG_CONTEXT,
+            ],
+            description="Ultra-fast small model",
+            recommended_for=["quick tasks", "prototyping"],
+            released="2024-07",
+        ),
+    },
+    # =========================================================================
+    # OPENROUTER
+    # =========================================================================
+    "openrouter": {
+        "anthropic/claude-sonnet-4": ModelInfo(
+            id="anthropic/claude-sonnet-4",
+            name="Claude Sonnet 4 (via OpenRouter)",
+            provider="openrouter",
+            input_price=3.0,
+            output_price=15.0,
+            context_window=200000,
+            max_output=8192,
+            capabilities=[
+                ModelCapability.TOOLS,
+                ModelCapability.VISION,
+                ModelCapability.STREAMING,
+                ModelCapability.CODE,
+                ModelCapability.LONG_CONTEXT,
+            ],
+            description="Claude via OpenRouter",
+            recommended_for=["coding", "general"],
+        ),
+        "openai/gpt-4o": ModelInfo(
+            id="openai/gpt-4o",
+            name="GPT-4o (via OpenRouter)",
+            provider="openrouter",
+            input_price=2.50,
+            output_price=10.0,
+            context_window=128000,
+            max_output=16384,
+            capabilities=[
+                ModelCapability.TOOLS,
+                ModelCapability.VISION,
+                ModelCapability.STREAMING,
+                ModelCapability.CODE,
+                ModelCapability.LONG_CONTEXT,
+            ],
+            description="GPT-4o via OpenRouter",
+            recommended_for=["general", "coding"],
+        ),
+        "google/gemini-2.0-flash": ModelInfo(
+            id="google/gemini-2.0-flash",
+            name="Gemini 2.0 Flash (via OpenRouter)",
+            provider="openrouter",
+            input_price=0.10,
+            output_price=0.40,
+            context_window=1000000,
+            max_output=8192,
+            capabilities=[
+                ModelCapability.TOOLS,
+                ModelCapability.VISION,
+                ModelCapability.STREAMING,
+                ModelCapability.LONG_CONTEXT,
+            ],
+            description="Gemini via OpenRouter",
+            recommended_for=["long context"],
+        ),
+    },
+    # =========================================================================
+    # OLLAMA (Local - Free)
+    # =========================================================================
+    "ollama": {
+        "llama3.2:3b": ModelInfo(
+            id="llama3.2:3b",
+            name="Llama 3.2 3B",
+            provider="ollama",
+            input_price=0.0,
+            output_price=0.0,
+            context_window=128000,
+            max_output=4096,
+            capabilities=[
+                ModelCapability.STREAMING,
+            ],
+            description="Small local model",
+            recommended_for=["quick local tasks"],
+        ),
+        "qwen2.5-coder:7b": ModelInfo(
+            id="qwen2.5-coder:7b",
+            name="Qwen 2.5 Coder 7B",
+            provider="ollama",
+            input_price=0.0,
+            output_price=0.0,
+            context_window=32000,
+            max_output=4096,
+            capabilities=[
+                ModelCapability.STREAMING,
+                ModelCapability.CODE,
+            ],
+            description="Local coding model",
+            recommended_for=["local coding"],
+        ),
+        "qwen2.5-coder:32b": ModelInfo(
+            id="qwen2.5-coder:32b",
+            name="Qwen 2.5 Coder 32B",
+            provider="ollama",
+            input_price=0.0,
+            output_price=0.0,
+            context_window=32000,
+            max_output=4096,
+            capabilities=[
+                ModelCapability.STREAMING,
+                ModelCapability.CODE,
+            ],
+            description="Larger local coding model",
+            recommended_for=["local coding", "complex tasks"],
+        ),
+    },
+}
+
+
+# ============================================================================
+# LIVE DATA INTEGRATION (models.dev)
+# ============================================================================
+
+# Flag to track if live data is available
+_live_models: Optional[Dict[str, Dict[str, ModelInfo]]] = None
+_use_live_data: bool = False
+
+
+def set_live_models(models: Dict[str, Dict[str, ModelInfo]]) -> None:
+    """
+    Set live model data from models.dev.
+
+    Called by the models_dev module after fetching fresh data.
+    """
+    global _live_models, _use_live_data
+    _live_models = models
+    _use_live_data = True
+
+
+def get_effective_models() -> Dict[str, Dict[str, ModelInfo]]:
+    """
+    Get the effective models database.
+
+    Returns live data if available, otherwise falls back to hardcoded MODELS.
+    """
+    if _use_live_data and _live_models:
+        # Merge: live data takes precedence, but keep hardcoded for missing providers
+        merged = MODELS.copy()
+        for provider_id, models in _live_models.items():
+            if provider_id in merged:
+                # Merge models within provider (live takes precedence)
+                merged[provider_id] = {**merged[provider_id], **models}
+            else:
+                merged[provider_id] = models
+        return merged
+    return MODELS
+
+
+# ============================================================================
+# HELPER FUNCTIONS
+# ============================================================================
+
+
+def get_model_info(provider_id: str, model_id: str) -> Optional[ModelInfo]:
+    """Get model info by provider and model ID."""
+    models = get_effective_models()
+    provider_models = models.get(provider_id, {})
+
+    # Try exact match first
+    if model_id in provider_models:
+        return provider_models[model_id]
+
+    # Try fuzzy match (partial ID match)
+    for mid, info in provider_models.items():
+        if model_id in mid or mid in model_id:
+            return info
+
+    return None
+
+
+def get_models_for_provider(provider_id: str) -> Dict[str, ModelInfo]:
+    """Get all models for a provider, filtered to ensure they belong to that provider."""
+    models = get_effective_models().get(provider_id, {})
+
+    # Filter models to ensure they actually belong to this provider
+    # This prevents models from other providers (e.g., GPT OSS) from appearing in Google's list
+    filtered = {}
+    for model_id, model_info in models.items():
+        # Ensure the model's provider field matches the requested provider
+        if model_info.provider == provider_id:
+            filtered[model_id] = model_info
+        # Special case for Google: only include Gemini models
+        elif provider_id == "google":
+            model_id_lower = model_id.lower()
+            model_name_lower = model_info.name.lower() if model_info.name else ""
+            # Only include if it's clearly a Gemini model
+            if (
+                ("gemini" in model_id_lower or "gemini" in model_name_lower)
+                and "gpt" not in model_id_lower
+                and "gpt" not in model_name_lower
+            ):
+                filtered[model_id] = model_info
+
+    return filtered
+
+
+def get_all_models() -> List[ModelInfo]:
+    """Get all models across all providers."""
+    all_models = []
+    for provider_models in get_effective_models().values():
+        all_models.extend(provider_models.values())
+    return all_models
+
+
+def get_all_providers() -> List[str]:
+    """Get all available provider IDs."""
+    return list(get_effective_models().keys())
+
+
+def get_cheapest_models(limit: int = 5) -> List[ModelInfo]:
+    """Get the cheapest models by input price."""
+    all_models = get_all_models()
+    sorted_models = sorted(all_models, key=lambda m: m.input_price)
+    return sorted_models[:limit]
+
+
+def get_models_with_capability(capability: ModelCapability) -> List[ModelInfo]:
+    """Get models that have a specific capability."""
+    return [m for m in get_all_models() if capability in m.capabilities]
+
+
+def get_recommended_for_coding() -> List[ModelInfo]:
+    """Get models recommended for coding."""
+    return [m for m in get_all_models() if "coding" in m.recommended_for or m.is_code_optimized]
+
+
+def search_models(query: str, limit: int = 20) -> List[ModelInfo]:
+    """
+    Search models by name, ID, or provider.
+
+    Args:
+        query: Search string (case-insensitive)
+        limit: Maximum results to return
+    """
+    query_lower = query.lower()
+    results = []
+
+    for model in get_all_models():
+        score = 0
+        # Exact ID match
+        if query_lower == model.id.lower():
+            score = 100
+        # ID contains query
+        elif query_lower in model.id.lower():
+            score = 80
+        # Name contains query
+        elif query_lower in model.name.lower():
+            score = 60
+        # Provider contains query
+        elif query_lower in model.provider.lower():
+            score = 40
+        # Description contains query
+        elif model.description and query_lower in model.description.lower():
+            score = 20
+
+        if score > 0:
+            results.append((score, model))
+
+    # Sort by score descending, then by name
+    results.sort(key=lambda x: (-x[0], x[1].name))
+    return [model for _, model in results[:limit]]
+
+
+def estimate_session_cost(
+    provider_id: str,
+    model_id: str,
+    input_tokens: int,
+    output_tokens: int,
+) -> Tuple[float, str]:
+    """
+    Estimate cost for a session.
+
+    Returns:
+        (cost, formatted_string)
+    """
+    info = get_model_info(provider_id, model_id)
+    if info:
+        cost = info.estimate_cost(input_tokens, output_tokens)
+        return cost, f"${cost:.4f}"
+    return 0.0, "Unknown"
+
+
+def is_using_live_data() -> bool:
+    """Check if live models.dev data is being used."""
+    return _use_live_data and _live_models is not None
+
+
+def get_data_source() -> str:
+    """Get a description of the current data source."""
+    if is_using_live_data():
+        return "models.dev (live)"
+    return "built-in (offline)"
+
+
+# ============================================================================
+# EXPORTS
+# ============================================================================
+
+__all__ = [
+    "ModelInfo",
+    "ModelCapability",
+    "MODELS",
+    "get_model_info",
+    "get_models_for_provider",
+    "get_all_models",
+    "get_all_providers",
+    "get_cheapest_models",
+    "get_models_with_capability",
+    "get_recommended_for_coding",
+    "search_models",
+    "estimate_session_cost",
+    "set_live_models",
+    "get_effective_models",
+    "is_using_live_data",
+    "get_data_source",
+]