LiteLLM: Authenticated SSRF and provider-credential exfiltration via unvalidated request-body routing parameters
BerriAI litellm
Affected versions
| Product | Affected | Fixed in |
|---|---|---|
| litellm PyPI |
|
BerriAI litellm
| Product | Affected | Fixed in |
|---|---|---|
| litellm PyPI |
|
LiteLLM proxy request validation does not fully check routing and credential parameters across all request fields. An authenticated proxy user can redirect outbound provider calls to a destination they control, exposing configured provider credentials, or make server-side requests to internal services reachable by the proxy.
Operators of the PyPI litellm package should check deployments on versions <1.88.6, >=1.89.0 <1.89.7, >=1.90.0 <1.90.7, >=1.91.0 <1.91.5, >=1.92.0 <1.92.2, >=1.93.0 <1.93.2, >=1.94.0 <1.94.3, >=1.95.0 <1.95.1, or >=1.96.0 <1.96.2.
|
LiteLLM is a proxy server (AI Gateway) to call LLM APIs in OpenAI (or native) format. Prior to versions 1.88.6 and 1.96.2, any authenticated LiteLLM proxy user could redirect an outbound provider call to a destination the user controls and cause the proxy to send its configured provider credentials to that destination. Request validation in litellm/proxy/auth/auth_utils.py, litellm/proxy/common_request_processing.py, litellm/proxy/health_endpoints/_health_endpoints.py, litellm/proxy/image_endpoints/endpoints.py, and litellm/proxy/litellm_pre_call_utils.py used incomplete checks that did not cover every sensitive parameter or inspect equivalent values across nested request fields, path values, and bracket-notation form data. Routing and credential parameters including api_base, base_url, model_list, fallbacks, and litellm_credential_name could therefore be applied without clearing the operator's stored key, exposing upstream provider credentials and other configured secrets and permitting server-side requests to internal services reachable by the proxy. This issue is fixed in versions 1.88.6 and 1.96.2.
From the CVE record · GitHub_M