docs: update all model references from Claude 4.5 to Claude 4.6 (anthropics#375)

* docs: update all model references from Claude 4.5 to Claude 4.6

- Update model IDs across all notebooks, scripts, and configs
- Add Bedrock model ID conventions and exclude from dated-model checks
- Add validation rules to prevent dated model IDs

* fix: update prose references from Claude Sonnet 4.5/Opus 4.5 to 4.6

- memory_cookbook.ipynb: Update 'Claude Sonnet 4.5' to 'Claude Sonnet 4.6'
- 01_The_chief_of_staff_agent.ipynb: Update 'Opus 4.5' to 'Opus 4.6'
- registry.yaml: Update title to 'Claude Sonnet 4.6'

Author: PedramNavid

.claude/skills/cookbook-audit/SKILL.md

@@ -123,7 +123,8 @@ Use this to ensure comprehensive coverage:
 **Technical Requirements**
 - [ ] Executable without modification (except API keys)
 - [ ] Uses non-deprecated API patterns
-- [ ] Uses valid model names (claude-sonnet-4-5, claude-haiku-4-5, claude-opus-4-5)
+- [ ] Uses valid model names (claude-sonnet-4-6, claude-haiku-4-5, claude-opus-4-6)
+- [ ] Uses non-dated model aliases (never dated IDs like claude-sonnet-4-6-20250514)
 - [ ] Model name defined as constant at top of notebook
 - [ ] Includes dependency specifications
 - [ ] Assigned to primary category

.claude/skills/cookbook-audit/style_guide.md

@@ -142,7 +142,7 @@ from dotenv import load_dotenv
 
 load_dotenv()
 
-MODEL = "claude-sonnet-4-5"
+MODEL = "claude-sonnet-4-6"
 
 client = anthropic.Anthropic()
 ~~~

.claude/skills/cookbook-audit/validate_notebook.py

@@ -321,20 +321,20 @@ def check_model_constant(self):
             if has_model_refs:
                 self.warnings.append(
                     "Model name should be defined as a constant at the top of the notebook "
-                    "(e.g., MODEL = 'claude-sonnet-4-5') to make future updates easier"
+                    "(e.g., MODEL = 'claude-sonnet-4-6') to make future updates easier"
                 )
 
     def check_deprecated_patterns(self):
         """Check for deprecated API patterns and invalid models."""
         # Valid models
-        valid_models = ["claude-sonnet-4-5", "claude-haiku-4-5", "claude-opus-4-5"]
+        valid_models = ["claude-sonnet-4-6", "claude-haiku-4-5", "claude-opus-4-6"]
 
         # Pattern to match model strings
         model_pattern = r'["\']claude-([a-z0-9\.-]+)["\']'
 
         # Deprecated models that should be migrated
         deprecated_models = {
-            "claude-opus-4-1": "claude-opus-4-5",
+            "claude-opus-4-1": "claude-opus-4-6",
         }
 
         deprecated_patterns = {
@@ -347,6 +347,17 @@ def check_deprecated_patterns(self):
 
             source = self.get_cell_source(cell)
 
+            # Check for dated API model IDs (exclude Bedrock IDs which require dates)
+            dated_pattern = r'["\']claude-\w+-[\d.]+-\d{8}["\']'
+            dated_matches = re.findall(dated_pattern, source)
+            for match in dated_matches:
+                # Skip Bedrock model IDs (anthropic.claude-*) which require dates
+                if f"anthropic.{match.strip(chr(34)).strip(chr(39))}" not in source:
+                    self.issues.append(
+                        f"Cell {i}: Dated model ID {match} found. "
+                        "Use non-dated aliases instead (e.g., claude-sonnet-4-6)"
+                    )
+
             # Check for invalid or deprecated models
             model_matches = re.findall(model_pattern, source)
             for match in model_matches:

.env.example

@@ -5,8 +5,8 @@
 ANTHROPIC_API_KEY=sk-ant-api03-...
 
 # Optional: Default model for testing (recommended for cost savings)
-# Note: claude-haiku-4-5-20251001 is the latest Haiku model as of October 2025
-CLAUDE_MODEL=claude-haiku-4-5-20251001
+# Note: claude-haiku-4-5 is the latest Haiku model as of October 2025
+CLAUDE_MODEL=claude-haiku-4-5
 
 # Optional: Test mode settings
 TEST_MODE=true

.github/workflows/notebook-tests.yml

@@ -146,7 +146,7 @@ jobs:
             - Explains how to fix common issues:
               - "Cells not executed": Run all cells from top to bottom before committing
               - "Execution order issues": Restart kernel and run all cells sequentially
-              - "Deprecated models": Update to current model versions (claude-sonnet-4-5-20250929, etc.)
+              - "Deprecated models": Update to current model versions (claude-sonnet-4-6, etc.)
               - "Hardcoded API keys": Use os.environ.get("ANTHROPIC_API_KEY") instead
             - Mentions they can test locally with: `make test-notebooks NOTEBOOK=path/to/notebook.ipynb`
             - Uses friendly, constructive language

CLAUDE.md

@@ -62,9 +62,16 @@ style: lint/format
 2. **Dependencies:** Use `uv add <package>` or `uv add --dev <package>`. Never edit pyproject.toml directly.
 
 3. **Models:** Use current Claude models. Check docs.anthropic.com for latest versions.
-   - Sonnet: `claude-sonnet-4-5-20250929`
-   - Haiku: `claude-haiku-4-5-20251001`
-   - Opus: `claude-opus-4-5-20251101`
+   - Sonnet: `claude-sonnet-4-6`
+   - Haiku: `claude-haiku-4-5`
+   - Opus: `claude-opus-4-6`
+   - **Never use dated model IDs** (e.g., `claude-sonnet-4-6-20250514`). Always use the non-dated alias.
+   - **Bedrock model IDs** follow a different format. Use the base Bedrock model ID from the docs:
+     - Opus 4.6: `anthropic.claude-opus-4-6-v1`
+     - Sonnet 4.5: `anthropic.claude-sonnet-4-5-20250929-v1:0`
+     - Haiku 4.5: `anthropic.claude-haiku-4-5-20251001-v1:0`
+     - Prepend `global.` for global endpoints (recommended): `global.anthropic.claude-opus-4-6-v1`
+     - Note: Bedrock models before Opus 4.6 require dated IDs in their Bedrock model ID.
 
 4. **Notebooks:**
    - Keep outputs in notebooks (intentional for demonstration)

CONTRIBUTING.md

@@ -118,7 +118,7 @@ If a hook fails, fix the issues and try committing again.
 
 2. **Use current Claude models**:
    - Use model aliases for better maintainability when available
-   - Latest Haiku model: `claude-haiku-4-5-20251001` (Haiku 4.5)
+   - Latest Haiku model: `claude-haiku-4-5` (Haiku 4.5)
    - Check current models at: https://docs.claude.com/en/docs/about-claude/models/overview
    - Claude will automatically validate model usage in PR reviews
 

capabilities/retrieval_augmented_generation/evaluation/eval_end_to_end.py

@@ -35,7 +35,7 @@ def evaluate_end_to_end(query, generated_answer, correct_answer):
     client = Anthropic(api_key=os.environ.get("ANTHROPIC_API_KEY"))
     try:
         response = client.messages.create(
-            model="claude-sonnet-4-5",
+            model="claude-sonnet-4-6",
             max_tokens=1500,
             messages=[
                 {"role": "user", "content": prompt},

capabilities/retrieval_augmented_generation/evaluation/promptfooconfig_end_to_end.yaml

@@ -13,7 +13,7 @@ providers:
       max_tokens: 2500
       temperature: 0
 
-  - id: anthropic:messages:claude-sonnet-4-5
+  - id: anthropic:messages:claude-sonnet-4-6
     label: "3.5 Sonnet: T-0.0"
     config:
       max_tokens: 2500

capabilities/retrieval_augmented_generation/evaluation/provider_retrieval.py

@@ -69,7 +69,7 @@ def _rerank_results(query: str, results: list[dict], k: int = 3) -> list[dict]:
     client = Anthropic(api_key=os.environ.get("ANTHROPIC_API_KEY"))
     try:
         response = client.messages.create(
-            model="claude-sonnet-4-5",
+            model="claude-sonnet-4-6",
             max_tokens=50,
             messages=[
                 {"role": "user", "content": prompt},