fix(llm): infer provider from naming patterns and sync anthropic models

[prvthmpcypher]

Summary

Fixes #6813.

Problem

1. In LLM._infer_provider_from_model, model provider resolution only checked hardcoded constant lists and defaulted directly to "openai" without calling _matches_provider_pattern(). This caused models like claude-sonnet-4-6 or custom Anthropic/Gemini models to incorrectly route to the OpenAI native client.
2. ANTHROPIC_MODELS and AnthropicModels literal in lib/crewai/src/crewai/llms/constants.py had drifted from lib/cli/src/crewai_cli/constants.py, missing claude-opus-4-6, claude-sonnet-4-6, claude-opus-4-7, claude-opus-4-8, claude-opus-5, claude-sonnet-5, claude-fable-5, and contained two invalid IDs (claude-4-sonnet-20250514, claude-4-opus-20250514).

Solution

• Updated LLM._infer_provider_from_model to fall through to _matches_provider_pattern for known providers before defaulting to "openai", as documented in its docstring.
• Synchronized ANTHROPIC_MODELS with the CLI model list and updated the AnthropicModels type alias.
• Added comprehensive unit tests in test_llm.py covering provider inference for Anthropic, Gemini, OpenAI, and custom prefix models.

Closes #6813

[coderabbitai]

📝 Walkthrough

Walkthrough

The change adds provider-pattern fallback to LLM provider inference, updates Anthropic model declarations with current Claude identifiers, removes deprecated identifiers, and adds parameterized provider-resolution tests.

Changes

Provider inference and Claude model updates

Layer / File(s)	Summary
Anthropic model declarations lib/crewai/src/crewai/llms/constants.py	The Anthropic literal type and runtime list add seven Claude identifiers and remove two deprecated identifiers.
Provider pattern fallback lib/crewai/src/crewai/llm.py	LLM._infer_provider_from_model uses provider-specific model-name patterns before defaulting unmatched models to OpenAI.
Provider inference validation lib/crewai/tests/test_llm.py	Parameterized tests verify Anthropic, Gemini, and OpenAI provider resolution for representative model names.

Sequence Diagram(s)

sequenceDiagram
  participant Caller
  participant LLM._infer_provider_from_model
  participant _matches_provider_pattern
  participant _get_native_provider
  Caller->>LLM._infer_provider_from_model: provide model name
  LLM._infer_provider_from_model->>_matches_provider_pattern: match unknown model name
  _matches_provider_pattern-->>LLM._infer_provider_from_model: return inferred provider
  LLM._infer_provider_from_model->>_get_native_provider: select inferred provider

Loading

Suggested reviewers: lorenzejay, greysonlalonde

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title clearly summarizes provider inference and Anthropic model synchronization, which are the primary changes.
Description check	✅ Passed	The description directly explains the provider inference fix, model-list synchronization, tests, and linked issue.
Linked Issues check	✅ Passed	The changes satisfy issue #6813 by adding pattern inference, synchronizing Anthropic models, removing invalid IDs, and adding tests.
Out of Scope Changes check	✅ Passed	All changes support issue #6813 and its stated objectives; no unrelated code changes are identified.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}

Warning

⚠️ This pull request shows signs of AI-generated slop (redundant_comments). It has been flagged by CodeRabbit slop detection and should be reviewed carefully.

[coderabbitai]

Actionable comments posted: 1

🧹 Nitpick comments (1)

lib/crewai/tests/test_llm.py (1)

1212-1229: 🎯 Functional Correctness | 🔵 Trivial | ⚡ Quick win

Exercise the new OpenAI pattern fallback.

The current cases do not include a clearly unlisted OpenAI model. They can pass through the hardcoded OPENAI_MODELS lookup without testing the new pattern branch. Add unlisted GPT identifiers with and without a dot, plus one unmatched model for the documented OpenAI default.

As per coding guidelines, **/*test*.py must write unit tests for new functionality, focusing on behavior rather than implementation details.

Proposed regression cases

         ("o3-mini", "openai"),
+        ("gpt-future-6", "openai"),
+        ("gpt-future-6.1", "openai"),
+        ("custom-model-v1", "openai"),

🤖 Prompt for AI Agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

In `@lib/crewai/tests/test_llm.py` around lines 1212 - 1229, Add cases to
test_infer_provider_from_model for unlisted GPT identifiers both with and
without a dot, ensuring they resolve to "openai" through the fallback, and add
one unmatched model that verifies the documented default also resolves to
"openai".

Source: Coding guidelines

🤖 Prompt for all review comments with AI agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

Inline comments:
In `@lib/crewai/src/crewai/llm.py`:
- Around line 662-664: Update the provider iteration in LLM provider resolution
so OpenAI-specific matching takes precedence over Azure and Bedrock for bare or
dotted GPT/o-series model names. Ensure generic gpt-* and o* models resolve to
"openai", while preserving explicit Azure and Bedrock matches for models
uniquely identified by those providers.

---

Nitpick comments:
In `@lib/crewai/tests/test_llm.py`:
- Around line 1212-1229: Add cases to test_infer_provider_from_model for
unlisted GPT identifiers both with and without a dot, ensuring they resolve to
"openai" through the fallback, and add one unmatched model that verifies the
documented default also resolves to "openai".

🪄 Autofix

Fix all unresolved CodeRabbit comments on this PR:

• Push a commit to this branch (recommended)
• Create a new PR with the fixes

---

ℹ️ Review info

⚙️ Run configuration

Configuration used: Organization UI

Review profile: CHILL

Plan: Pro Plus

Run ID: 2c39e75d-de37-49b7-9a0f-3165754fa830

📥 Commits

Reviewing files that changed from the base of the PR and between 094b94e and 19a5b13.

📒 Files selected for processing (3)

• lib/crewai/src/crewai/llm.py
• lib/crewai/src/crewai/llms/constants.py
• lib/crewai/tests/test_llm.py

[coderabbitai]

🎯 Functional Correctness | 🟠 Major | ⚡ Quick win

Evaluate specific provider patterns before broad matches.

For an unlisted gpt-* model, the Azure pattern matches before the OpenAI pattern. gpt-future-6 therefore resolves to "azure". A dotted unlisted GPT model resolves to "bedrock" because the Bedrock pattern accepts any model containing "." and runs first. LLM.__new__ then selects the wrong native provider.

Evaluate OpenAI before Azure and Bedrock for bare model names. Reserve generic gpt-* and o* matching for OpenAI, or make the Azure and Bedrock patterns mutually exclusive.

Proposed ordering fix

-        for provider in ["anthropic", "gemini", "bedrock", "azure", "openai"]:
+        for provider in ["anthropic", "gemini", "openai", "azure", "bedrock"]:

📝 Committable suggestion

‼️ IMPORTANT
Carefully review the code before committing. Ensure that it accurately replaces the highlighted code, contains no missing lines, and has no issues with indentation. Thoroughly test & benchmark the code to ensure it meets the requirements.

Suggested change

	for provider in ["anthropic", "gemini", "bedrock", "azure", "openai"]:
	if cls._matches_provider_pattern(model, provider):
	return provider
	for provider in ["anthropic", "gemini", "openai", "azure", "bedrock"]:
	if cls._matches_provider_pattern(model, provider):
	return provider

🤖 Prompt for AI Agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

In `@lib/crewai/src/crewai/llm.py` around lines 662 - 664, Update the provider
iteration in LLM provider resolution so OpenAI-specific matching takes
precedence over Azure and Bedrock for bare or dotted GPT/o-series model names.
Ensure generic gpt-* and o* models resolve to "openai", while preserving
explicit Azure and Bedrock matches for models uniquely identified by those
providers.