🔄 Upstream Sync: LiteLLM v1.83.3-stable

.circleci/requirements.txt

@@ -7,14 +7,15 @@ cohere
 redis==5.2.1
 redisvl==0.4.1
 anthropic
-orjson==3.10.12 # fast /embedding responses
-pydantic==2.11.0
-google-cloud-aiplatform==1.43.0
+orjson==3.10.15 # fast /embedding responses
+pydantic==2.12.5
+google-cloud-aiplatform==1.133.0
 google-cloud-iam==2.19.1
 fastapi-sso==0.16.0 
 uvloop==0.21.0
-mcp==1.25.0    # for MCP server  
+mcp==1.26.0    # for MCP server
 semantic_router==0.1.10 # for auto-routing with litellm
-fastuuid==0.12.0
+fastuuid==0.14.0
 responses==0.25.7  # for proxy client tests
-pytest-retry==1.6.3  # for automatic test retries
+pytest-retry==1.6.3  # for automatic test retries
+litellm-proxy-extras  # for prisma migrations

.devcontainer/post-create.sh

@@ -12,6 +12,6 @@ echo "[post-create] Generating Prisma client"
 poetry run prisma generate
 
 echo "[post-create] Installing npm dependencies"
-cd ui/litellm-dashboard && npm install --no-audit --no-fund
+cd ui/litellm-dashboard && npm ci
 
 echo "[post-create] Done"

.dockerignore

@@ -48,7 +48,7 @@ dist/
 build/
 *.egg-info/
 .DS_Store
-node_modules/
+**/node_modules
 *.log
 .env
 .env.local

.github/ISSUE_TEMPLATE/config.yml

@@ -1,7 +1,7 @@
 blank_issues_enabled: true
 contact_links:
   - name: Schedule Demo
-    url: https://calendly.com/d/4mp-gd3-k5k/litellm-1-1-onboarding-chat
+    url: https://enterprise.litellm.ai/demo
     about: Speak directly with Krrish and Ishaan, the founders, to discuss issues, share feedback, or explore improvements for LiteLLM
   - name: Discord
     url: https://discord.com/invite/wuPM9dRgDw

.github/actions/helm-oci-chart-releaser/action.yml

@@ -40,38 +40,55 @@ outputs:
 runs:
   using: composite
   steps:
+    - name: Helm | Setup
+      uses: azure/setup-helm@1a275c3b69536ee54be43f2070a358922e12c8d4 # v4.3.1
+      with:
+        version: v3.20.0
+
     - name: Helm | Login
       shell: bash
-      run: echo ${{ inputs.registry_password }} | helm registry login -u ${{ inputs.registry_username }} --password-stdin ${{ inputs.registry }}
       env:
-        HELM_EXPERIMENTAL_OCI: '1'
-
+        REGISTRY_PASSWORD: ${{ inputs.registry_password }}
+        REGISTRY_USERNAME: ${{ inputs.registry_username }}
+        REGISTRY: ${{ inputs.registry }}
+      run: echo "$REGISTRY_PASSWORD" | helm registry login -u "$REGISTRY_USERNAME" --password-stdin "$REGISTRY"
+
     - name: Helm | Dependency
       if: inputs.update_dependencies == 'true'
       shell: bash
-      run: helm dependency update ${{ inputs.path == null && format('{0}/{1}', 'charts', inputs.name) || inputs.path }}
       env:
-        HELM_EXPERIMENTAL_OCI: '1'
+        CHART_PATH: ${{ inputs.path == null && format('{0}/{1}', 'charts', inputs.name) || inputs.path }}
+      run: helm dependency update "$CHART_PATH"
 
     - name: Helm | Package
       shell: bash
-      run: helm package ${{ inputs.path == null && format('{0}/{1}', 'charts', inputs.name) || inputs.path }} --version ${{ inputs.tag }} --app-version ${{ inputs.app_version }}
       env:
-        HELM_EXPERIMENTAL_OCI: '1'
+        CHART_PATH: ${{ inputs.path == null && format('{0}/{1}', 'charts', inputs.name) || inputs.path }}
+        TAG: ${{ inputs.tag }}
+        APP_VERSION: ${{ inputs.app_version }}
+      run: helm package "$CHART_PATH" --version "$TAG" --app-version "$APP_VERSION"
 
     - name: Helm | Push
       shell: bash
-      run: helm push ${{ inputs.name }}-${{ inputs.tag }}.tgz oci://${{ inputs.registry }}/${{ inputs.repository }}
       env:
-        HELM_EXPERIMENTAL_OCI: '1'
+        NAME: ${{ inputs.name }}
+        TAG: ${{ inputs.tag }}
+        REGISTRY: ${{ inputs.registry }}
+        REPOSITORY: ${{ inputs.repository }}
+      run: helm push "${NAME}-${TAG}.tgz" "oci://${REGISTRY}/${REPOSITORY}"
 
     - name: Helm | Logout
       shell: bash
-      run: helm registry logout ${{ inputs.registry }}
       env:
-        HELM_EXPERIMENTAL_OCI: '1'
+        REGISTRY: ${{ inputs.registry }}
+      run: helm registry logout "$REGISTRY"
 
     - name: Helm | Output
       id: output
       shell: bash
-      run: echo "image=${{ inputs.registry }}/${{ inputs.repository }}/${{ inputs.name }}:${{ inputs.tag }}" >> $GITHUB_OUTPUT
+      env:
+        REGISTRY: ${{ inputs.registry }}
+        REPOSITORY: ${{ inputs.repository }}
+        NAME: ${{ inputs.name }}
+        TAG: ${{ inputs.tag }}
+      run: echo "image=${REGISTRY}/${REPOSITORY}/${NAME}:${TAG}" >> $GITHUB_OUTPUT

.github/codeql/codeql-config.yml

@@ -0,0 +1,21 @@
+name: "LiteLLM CodeQL config"
+
+queries:
+  - uses: security-and-quality
+
+# Known OOM queries on large Python codebases:
+# CodeQL builds a full data flow graph in memory. These two queries trace
+# sensitive data through every log call / regex pattern, causing combinatorial
+# path explosion on codebases with extensive logging like LiteLLM (>2 GiB
+# result sets). This is a known CodeQL scaling limitation, not a code issue.
+# Re-test periodically as CodeQL improves or the codebase refactors logging.
+query-filters:
+  - exclude:
+      id: py/clear-text-logging-sensitive-data  # CWE-312
+  - exclude:
+      id: py/polynomial-redos                   # CWE-730
+
+paths-ignore:
+  - tests
+  - docs
+  - "**/*.md"

.github/dependabot.yaml

@@ -4,6 +4,9 @@ updates:
     directory: "/"
     schedule:
       interval: "daily"
+    cooldown:
+      default-days: 7
+      semver-major-days: 14
     groups:
       github-actions:
         patterns:

.github/observatory/litellm_config.yaml

@@ -0,0 +1,19 @@
+# LiteLLM Observatory Test Configuration
+# This config is used by CI to spin up a temporary LiteLLM instance
+# for running observatory tests against RC/stable releases.
+#
+# Add model definitions for the providers you want to test.
+# Provider API keys are injected via environment variables in CI.
+
+model_list:
+  - model_name: gpt-4o
+    litellm_params:
+      model: azure/gpt-4o
+      api_key: os.environ/AZURE_API_KEY
+      api_base: os.environ/AZURE_API_BASE
+
+  - model_name: gpt-4o-mini
+    litellm_params:
+      model: azure/gpt-4o-mini
+      api_key: os.environ/AZURE_API_KEY
+      api_base: os.environ/AZURE_API_BASE

.github/pull_request_template.md

@@ -6,9 +6,14 @@
 
 **Please complete all items before asking a LiteLLM maintainer to review your PR**
 
-- [ ] I have Added testing in the [`tests/litellm/`](https://github.com/BerriAI/litellm/tree/main/tests/litellm) directory, **Adding at least 1 test is a hard requirement** - [see details](https://docs.litellm.ai/docs/extras/contributing_code)
+- [ ] I have Added testing in the [`tests/test_litellm/`](https://github.com/BerriAI/litellm/tree/main/tests/test_litellm) directory, **Adding at least 1 test is a hard requirement** - [see details](https://docs.litellm.ai/docs/extras/contributing_code)
 - [ ] My PR passes all unit tests on [`make test-unit`](https://docs.litellm.ai/docs/extras/contributing_code)
 - [ ] My PR's scope is as isolated as possible, it only solves 1 specific problem
+- [ ] I have requested a Greptile review by commenting `@greptileai` and received a **Confidence Score of at least 4/5** before requesting a maintainer review
+
+## Delays in PR merge?
+
+If you're seeing a delay in your PR being merged, ping the LiteLLM Team on [Slack (#pr-review)](https://join.slack.com/t/litellmossslack/shared_invite/zt-3o7nkuyfr-p_kbNJj8taRfXGgQI1~YyA).
 
 ## CI (LiteLLM team)
 

.github/scripts/close_duplicate_issues.py

@@ -0,0 +1,208 @@
+#!/usr/bin/env python3
+"""
+Detect and close duplicate GitHub issues using title similarity.
+
+Modes:
+  --scan           Compare all open issues against each other (batch)
+  --issue-number N Check a single issue against older open issues
+
+Requires the `gh` CLI to be authenticated.
+"""
+
+import argparse
+import difflib
+import json
+import re
+import subprocess
+import sys
+
+
+def normalize_title(title: str) -> str:
+    """Strip common prefixes, lowercase, and collapse whitespace."""
+    title = re.sub(
+        r"^\[?(bug|feature request|enhancement|question|docs)[:\]]?\s*",
+        "",
+        title,
+        flags=re.IGNORECASE,
+    )
+    return " ".join(title.lower().split())
+
+
+def gh(*args: str) -> str:
+    """Run a gh CLI command and return stdout."""
+    result = subprocess.run(
+        ["gh", *args],
+        capture_output=True,
+        text=True,
+        check=True,
+    )
+    return result.stdout
+
+
+def fetch_open_issues(repo: str | None) -> list[dict]:
+    """Fetch all open issues (excluding PRs) via gh api --paginate."""
+    if repo:
+        endpoint = f"repos/{repo}/issues?state=open&per_page=100&sort=created&direction=asc"
+    else:
+        endpoint = "repos/{owner}/{repo}/issues?state=open&per_page=100&sort=created&direction=asc"
+    cmd = ["api", "--paginate", endpoint]
+
+    raw = gh(*cmd)
+    # gh --paginate concatenates JSON arrays, so we may get multiple arrays
+    issues = []
+    for line in raw.strip().splitlines():
+        line = line.strip()
+        if not line:
+            continue
+        parsed = json.loads(line)
+        if isinstance(parsed, list):
+            issues.extend(parsed)
+        else:
+            issues.append(parsed)
+
+    # Filter out pull requests (they also appear in the issues endpoint)
+    return [i for i in issues if "pull_request" not in i]
+
+
+def close_as_duplicate(
+    issue_number: int, duplicate_of: int, repo: str | None, dry_run: bool
+) -> None:
+    """Close an issue as duplicate of another, adding a comment and label."""
+    repo_args = ["--repo", repo] if repo else []
+
+    if dry_run:
+        print(f"  [DRY RUN] Would close #{issue_number} as duplicate of #{duplicate_of}")
+        return
+
+    # Add comment
+    comment_body = (
+        f"Closing as duplicate of #{duplicate_of}.\n\n"
+        "If you believe this is not a duplicate, please reopen and add context "
+        "explaining how this differs."
+    )
+    gh("issue", "comment", str(issue_number), "--body", comment_body, *repo_args)
+
+    # Add label
+    gh("issue", "edit", str(issue_number), "--add-label", "duplicate", *repo_args)
+
+    # Close with not_planned reason
+    gh(
+        "api",
+        f"repos/{repo or '{owner}/{repo}'}/issues/{issue_number}",
+        "-X",
+        "PATCH",
+        "-f",
+        "state=closed",
+        "-f",
+        "state_reason=not_planned",
+    )
+
+    print(f"  Closed #{issue_number} as duplicate of #{duplicate_of}")
+
+
+def find_duplicate(
+    issue: dict, candidates: list[dict], threshold: float
+) -> dict | None:
+    """Return the first candidate whose normalized title is above threshold."""
+    norm = normalize_title(issue["title"])
+    for candidate in candidates:
+        if candidate["number"] == issue["number"]:
+            continue
+        cand_norm = normalize_title(candidate["title"])
+        ratio = difflib.SequenceMatcher(None, norm, cand_norm).ratio()
+        if ratio >= threshold:
+            return candidate
+    return None
+
+
+def scan_all(issues: list[dict], threshold: float, repo: str | None, dry_run: bool) -> int:
+    """Compare every issue against all older issues. Returns count of duplicates found."""
+    # Sort oldest first
+    issues.sort(key=lambda i: i["number"])
+    closed_count = 0
+
+    for idx, issue in enumerate(issues):
+        older = issues[:idx]
+        if not older:
+            continue
+        dup = find_duplicate(issue, older, threshold)
+        if dup:
+            ratio = difflib.SequenceMatcher(
+                None,
+                normalize_title(issue["title"]),
+                normalize_title(dup["title"]),
+            ).ratio()
+            print(
+                f"#{issue['number']}: \"{issue['title']}\"\n"
+                f"  -> duplicate of #{dup['number']}: \"{dup['title']}\" "
+                f"({ratio:.0%} similar)"
+            )
+            close_as_duplicate(issue["number"], dup["number"], repo, dry_run)
+            closed_count += 1
+
+    return closed_count
+
+
+def check_single(
+    issue_number: int, issues: list[dict], threshold: float, repo: str | None, dry_run: bool
+) -> bool:
+    """Check a single issue against all older open issues. Returns True if duplicate found."""
+    target = None
+    for i in issues:
+        if i["number"] == issue_number:
+            target = i
+            break
+
+    if target is None:
+        print(f"Issue #{issue_number} not found among open issues.")
+        return False
+
+    older = [i for i in issues if i["number"] < issue_number]
+    dup = find_duplicate(target, older, threshold)
+    if dup:
+        ratio = difflib.SequenceMatcher(
+            None,
+            normalize_title(target["title"]),
+            normalize_title(dup["title"]),
+        ).ratio()
+        print(
+            f"#{target['number']}: \"{target['title']}\"\n"
+            f"  -> duplicate of #{dup['number']}: \"{dup['title']}\" "
+            f"({ratio:.0%} similar)"
+        )
+        close_as_duplicate(issue_number, dup["number"], repo, dry_run)
+        return True
+
+    print(f"#{issue_number}: no duplicate found above threshold {threshold}")
+    return False
+
+
+def main() -> None:
+    parser = argparse.ArgumentParser(description="Detect and close duplicate GitHub issues")
+    mode = parser.add_mutually_exclusive_group(required=True)
+    mode.add_argument("--scan", action="store_true", help="Scan all open issues")
+    mode.add_argument("--issue-number", type=int, help="Check a single issue number")
+    parser.add_argument("--threshold", type=float, default=0.85, help="Similarity threshold (0-1)")
+    parser.add_argument("--close", action="store_true", help="Actually close duplicates (default is dry-run)")
+    parser.add_argument("--repo", type=str, help="Repository (owner/repo). Auto-detected if omitted.")
+    args = parser.parse_args()
+
+    dry_run = not args.close
+
+    if dry_run:
+        print("=== DRY RUN MODE (pass --close to actually close issues) ===\n")
+
+    print("Fetching open issues...")
+    issues = fetch_open_issues(args.repo)
+    print(f"Found {len(issues)} open issues.\n")
+
+    if args.scan:
+        count = scan_all(issues, args.threshold, args.repo, dry_run)
+        print(f"\nTotal duplicates {'found' if dry_run else 'closed'}: {count}")
+    else:
+        found = check_single(args.issue_number, issues, args.threshold, args.repo, dry_run)
+        sys.exit(0 if found else 0)  # Always exit 0; finding no dup is not an error
+
+
+if __name__ == "__main__":
+    main()