docs: add context limit and strategy to LLM model configs API

- Add `context_limit` as required field for LLM model configurations
- Add `context_strategy` as optional field with three trimming strategies
- Update API documentation with detailed examples for GPT-4o and Claude 3.5 Sonnet
- Clarify that context limit is specified in K tokens (thousands of tokens)
- Update example payloads to reflect current model versions and new fields

scripts/show_model_context_limits.py

@@ -0,0 +1,157 @@
+#!/usr/bin/env python3
+"""
+Reference tool for LLM model context limits.
+
+This script provides reference context limit values for common LLM models.
+It helps users find the correct context_limit value when creating LLM model configurations.
+
+IMPORTANT:
+- context_limit unit is K tokens (1 K = 1000 tokens)
+- This tool only displays reference values. You MUST manually configure
+  context_limit when creating or updating LLM model configurations via API.
+
+Usage:
+    python scripts/show_model_context_limits.py
+
+For official documentation, always check:
+- OpenAI: https://platform.openai.com/docs/models
+- Anthropic: https://docs.anthropic.com/claude/docs/models-overview
+- Google: https://ai.google.dev/gemini-api/docs/models
+- DeepSeek: https://platform.deepseek.com/api-docs/
+"""
+
+import sys
+from pathlib import Path
+
+# Add parent directory to path
+sys.path.insert(0, str(Path(__file__).parent.parent))
+
+
+# Reference context limits (as of 2025)
+# Displayed in K tokens for easier configuration
+# Users should verify from official provider documentation
+MODEL_CONTEXT_LIMITS_K = {
+    # OpenAI Models
+    "gpt-4o": 128,
+    "gpt-4o-mini": 128,
+    "gpt-4-turbo": 128,
+    "gpt-4": 8,
+    "gpt-4-32k": 33,
+    "gpt-3.5-turbo": 17,
+    "gpt-3.5-turbo-16k": 17,
+
+    # Anthropic Models
+    "claude-3-5-sonnet-20241022": 200,
+    "claude-3-5-sonnet-20240620": 200,
+    "claude-3-opus-20240229": 200,
+    "claude-3-sonnet-20240229": 200,
+    "claude-3-haiku-20240307": 200,
+
+    # Google Models
+    "gemini-2.0-flash-exp": 1000,
+    "gemini-1.5-pro": 2800,
+    "gemini-1.5-flash": 2800,
+    "gemini-pro": 92,
+
+    # DeepSeek Models
+    "deepseek-chat": 128,
+    "deepseek-coder": 128,
+
+    # xAI Models
+    "grok-beta": 128,
+}
+
+
+def find_context_limit_for_model(model_name: str) -> int | None:
+    """Find the context limit for a given model name (in K tokens)."""
+    model_lower = model_name.lower().strip()
+
+    # Try exact match
+    if model_lower in MODEL_CONTEXT_LIMITS_K:
+        return MODEL_CONTEXT_LIMITS_K[model_lower]
+
+    # Try prefix match
+    for key, limit in MODEL_CONTEXT_LIMITS_K.items():
+        if model_lower.startswith(key.lower()):
+            return limit
+
+    return None
+
+
+def main():
+    print("=" * 70)
+    print("LLM Model Context Limits Reference Tool")
+    print("=" * 70)
+    print()
+    print("This tool displays reference context limit values for common LLM models.")
+    print("Please verify from official provider documentation before configuring.")
+    print()
+    print("IMPORTANT: context_limit unit is K tokens (1 K = 1,000 tokens)")
+    print()
+    print("Official Documentation:")
+    print("  - OpenAI: https://platform.openai.com/docs/models")
+    print("  - Anthropic: https://docs.anthropic.com/claude/docs/models-overview")
+    print("  - Google: https://ai.google.dev/gemini-api/docs/models")
+    print("  - DeepSeek: https://platform.deepseek.com/api-docs/")
+    print()
+
+    # Display all reference values
+    print("=" * 70)
+    print("Reference Context Limits (in K tokens)")
+    print("=" * 70)
+    print()
+
+    # Group by provider
+    providers = {
+        "OpenAI": ["gpt-4o", "gpt-4o-mini", "gpt-4-turbo", "gpt-4", "gpt-4-32k", "gpt-3.5-turbo", "gpt-3.5-turbo-16k"],
+        "Anthropic": ["claude-3-5-sonnet-20241022", "claude-3-5-sonnet-20240620", "claude-3-opus-20240229", "claude-3-sonnet-20240229", "claude-3-haiku-20240307"],
+        "Google": ["gemini-2.0-flash-exp", "gemini-1.5-pro", "gemini-1.5-flash", "gemini-pro"],
+        "DeepSeek": ["deepseek-chat", "deepseek-coder"],
+        "xAI": ["grok-beta"],
+    }
+
+    for provider, models in providers.items():
+        print(f"\n{provider}:")
+        for model in models:
+            if model in MODEL_CONTEXT_LIMITS_K:
+                limit_k = MODEL_CONTEXT_LIMITS_K[model]
+                limit_actual = limit_k * 1000
+                print(f"  {model:40s} → {limit_k:4d}K (= {limit_actual:,} tokens)")
+
+    print()
+    print("=" * 70)
+    print()
+    print("Conversion Examples:")
+    print()
+    print("  Official Documentation: 128,000 tokens")
+    print("  ↓")
+    print("  API Configuration: \"context_limit\": 128")
+    print()
+    print("  Official Documentation: 200,000 tokens")
+    print("  ↓")
+    print("  API Configuration: \"context_limit\": 200")
+    print()
+    print("  Official Documentation: 2,800,000 tokens")
+    print("  ↓")
+    print("  API Configuration: \"context_limit\": 2800")
+    print()
+    print("=" * 70)
+    print()
+    print("Usage Example:")
+    print()
+    print("When creating a model configuration, specify context_limit in K:")
+    print()
+    print('  POST /v3/users/{user_id}/llm-model-configs')
+    print('  {')
+    print('    "name": "GPT-4o Configuration",')
+    print('    "provider": "openai",')
+    print('    "model": "gpt-4o",')
+    print('    "context_limit": 128,  // ← Required: 128K = 128,000 tokens')
+    print('    "context_strategy": "balanced"')
+    print('  }')
+    print()
+    print("=" * 70)
+
+
+if __name__ == "__main__":
+    main()