Update LiteLLM models

[bubacoder]

Summary by CodeRabbit

• Documentation

  • Added a comprehensive procedural guide for updating LiteLLM model configurations, including lookup steps, decision rules, update/verification commands, and provider model ID references.

• Updates

  • Rolled model registry to newer provider IDs: Anthropic, OpenAI/OpenRouter, Meta/LLama family, Deepseek and others—several models updated, replaced, or removed and fallback mappings adjusted.

[coderabbitai]

Walkthrough

Adds a new procedural doc for updating LiteLLM provider model IDs and updates the LiteLLM config with multiple provider model replacements and fallback adjustments (Anthropic Claude variants, OpenRouter/OpenAI gpt-4.1 changes, llama/deepseek replacements, and minor restructuring).

Changes

Cohort / File(s)	Summary
Documentation \.claude/commands/update-litellm-models.md	Adds a step-by-step guide to locate, evaluate, and apply latest provider model IDs to the LiteLLM YAML, including decision rules, CLI commands to write updates, verification steps, and ID-format references.
Model configuration docker/ai/litellm/config/config.yaml	Replaced/updated multiple provider model entries: Anthropic Claude updates (e.g., claude-sonnet-4-0 → claude-sonnet-4-6, claude-opus-4-6, claude-haiku-4-5-20251001), OpenRouter OpenAI adjustments (gpt-4o → gpt-4.1, gpt-4o-mini → gpt-4.1-mini), llama/deepseek replacements (llama-3.3-70b-instruct → llama-4-maverick, deepseek-r1-distill-llama-70b → deepseek-v3.2), and updated gpt-5 fallback mappings and related comments/structure.

Estimated code review effort

🎯 3 (Moderate) | ⏱️ ~20 minutes

Possibly related PRs

• Upgrade LiteLLM, add new models #237: Overlapping updates to the same LiteLLM config model entries (Anthropic/OpenRouter model ID changes).
• Update Anthropic models #213: Related Anthropic Claude model entry updates and litellm model string adjustments in the same config.
• Add new AI models, update docs #137: Changes to LiteLLM configuration organization and local/model entry rearrangements that overlap with these edits.

🚥 Pre-merge checks | ✅ 3

✅ Passed checks (3 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title 'Update LiteLLM models' directly and clearly summarizes the main change: updating LiteLLM model configurations across multiple providers in the config file.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Post copyable unit tests in a comment
• Commit unit tests in branch feature/models

---

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 and usage tips.}

[coderabbitai]

Actionable comments posted: 1

🤖 Prompt for all review comments with AI agents

Verify each finding against the current code and only fix it if needed.

Inline comments:
In @.claude/commands/update-litellm-models.md:
- Around line 48-50: The command currently writes to a machine-specific absolute
path (/mnt/local/storage/repos/infra/docker/ai/litellm/config/config.yaml);
change it to a repo-relative path (for example
infra/docker/ai/litellm/config/config.yaml or prefixed with $(git rev-parse
--show-toplevel)/infra/... or an environment variable like $REPO_ROOT) so the
tee invocation is portable across machines; update the tee target in the heredoc
invocation accordingly (replace the absolute path with the chosen repo-relative
or env-based path while keeping the heredoc content and EOF markers unchanged).

---

ℹ️ Review info

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Pro

📥 Commits

Reviewing files that changed from the base of the PR and between c4ef912 and 64afd4d.

📒 Files selected for processing (2)

• .claude/commands/update-litellm-models.md
• docker/ai/litellm/config/config.yaml

[coderabbitai]

🧹 Nitpick comments (2)

.claude/commands/update-litellm-models.md (1)

55-60: Verification step may show stale config if container isn't restarted.

Step 4 reads from the container's /app/config.yaml, but after updating the source file on disk, the container may still have the old config mounted or cached. Consider adding a restart step or verifying the source file directly:

Suggested enhancement

 ## Step 4: Verify

 Confirm the file was written correctly:
 ```bash
-docker exec litellm cat /app/config.yaml
+cat "$(git rev-parse --show-toplevel)/docker/ai/litellm/config/config.yaml"
+```
+
+Restart the container to apply changes:
+```bash
+docker restart litellm

</details>

<details>
<summary>🤖 Prompt for AI Agents</summary>

Verify each finding against the current code and only fix it if needed.

In @.claude/commands/update-litellm-models.md around lines 55 - 60, Update "Step
4: Verify" to avoid showing a stale in-container config: replace the direct
in-container cat command ("docker exec litellm cat /app/config.yaml") with a
verification of the host source file (the repo config path determined via git
root) and add a step to restart the container (docker restart litellm) so the
container picks up the updated config; ensure the doc shows verifying the host
config file first, then restarting the container, then (optionally) re-checking
the in-container /app/config.yaml.

</details>

</blockquote></details>
<details>
<summary>docker/ai/litellm/config/config.yaml (1)</summary><blockquote>

`23-36`: **Anthropic model updates look structurally correct.**

The claude-sonnet-4-6 and claude-opus-4-6 models are configured consistently. However, there's a naming mismatch for claude-haiku: the `model_name` is `claude-haiku-4-5` (alias) but the litellm model is `anthropic/claude-haiku-4-5-20251001` (full version ID). Consider aligning these for consistency.

<details>
<summary>Optional: Align model_name with the dated model ID</summary>

```diff
-  - model_name: claude-haiku-4-5
+  - model_name: claude-haiku-4-5-20251001
     litellm_params:
       model: anthropic/claude-haiku-4-5-20251001

🤖 Prompt for AI Agents

Verify each finding against the current code and only fix it if needed.

In `@docker/ai/litellm/config/config.yaml` around lines 23 - 36, The claude-haiku
entry has a naming mismatch: model_name is "claude-haiku-4-5" while
litellm_params.model is "anthropic/claude-haiku-4-5-20251001"; update one to
match the other for consistency (either change model_name to
"claude-haiku-4-5-20251001" or set litellm_params.model to
"anthropic/claude-haiku-4-5") so that the alias (model_name) and the actual
model identifier in litellm_params align; locate the claude-haiku-4-5 block in
the YAML and make the corresponding edit to model_name or litellm_params.model.

🤖 Prompt for all review comments with AI agents

Verify each finding against the current code and only fix it if needed.

Nitpick comments:
In @.claude/commands/update-litellm-models.md:
- Around line 55-60: Update "Step 4: Verify" to avoid showing a stale
in-container config: replace the direct in-container cat command ("docker exec
litellm cat /app/config.yaml") with a verification of the host source file (the
repo config path determined via git root) and add a step to restart the
container (docker restart litellm) so the container picks up the updated config;
ensure the doc shows verifying the host config file first, then restarting the
container, then (optionally) re-checking the in-container /app/config.yaml.

In `@docker/ai/litellm/config/config.yaml`:
- Around line 23-36: The claude-haiku entry has a naming mismatch: model_name is
"claude-haiku-4-5" while litellm_params.model is
"anthropic/claude-haiku-4-5-20251001"; update one to match the other for
consistency (either change model_name to "claude-haiku-4-5-20251001" or set
litellm_params.model to "anthropic/claude-haiku-4-5") so that the alias
(model_name) and the actual model identifier in litellm_params align; locate the
claude-haiku-4-5 block in the YAML and make the corresponding edit to model_name
or litellm_params.model.

---

ℹ️ Review info

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Pro

📥 Commits

Reviewing files that changed from the base of the PR and between 64afd4d and 36a6833.

📒 Files selected for processing (2)

• .claude/commands/update-litellm-models.md
• docker/ai/litellm/config/config.yaml