Veille CVE

Recherche « Litellm litellm »

10 CVE

Critique 3 Élevée 3 Moyenne 4 Faible 0 Non évaluée 0 Exploitées (KEV) Tri : plus récentes « Litellm litellm » ✕
  1. CVE-2026-114612 9.9 Critique

    BerriAI litellm

    LiteLLM is a proxy server (AI Gateway) to call LLM APIs in OpenAI (or native) format. From 1.87.0 until 1.100.4, 1.101.3, 1.102.2, 1.103.1, and 1.104.0-rc.2, the proxy reuses one cryptographic context for at-rest secrets and session tokens, allowing an authenticated internal user to have crafted metadata containing a forged administrator credential encrypted and returned by the API-key flow. Replaying that ciphertext as a bearer token causes the proxy to trust the forged proxy administrator identity, permitting administrative access and arbitrary commands through the MCP stdio endpoint. Versions 1.87.0 through 1.90.x require EXPERIMENTAL_UI_LOGIN to be enabled, while versions after 1.91.0 are exploitable in the default configuration. This issue was fixed in LiteLLM 1.100.4, 1.101.3, 1.102.2, 1.103.1, and 1.104.0-rc.2.

  2. CVE-2026-59970 7.7 Élevée

    TandoorRecipes recipes

    Tandoor Recipes is an application for managing recipes, planning meals, and building shopping lists. Prior to 2.6.10, passes an administrator-controlled AI provider URL directly to LiteLLM without applying the server's outbound-request SSRF checks when an authenticated user creates a Space, configures an AI provider with a URL targeting an internal address, and invokes an AI action. In cookbook/views/api.py, the affected AI actions call completion through TARGET/api/ai-import/ and api/ai-step-sort/ while bypassing safe_request. The attack works because the server connects to the supplied provider URL from each AI action and exposes connection success, refusal, or timeout as a blind network-reachability oracle. This can reach internal services, other hosts on the server network, and cloud metadata endpoints and map reachable internal hosts and ports. This issue is fixed in version 2.6.10.

  3. CVE-2026-108575 6.3 Moyenne

    BerriAI LiteLLM

    A vulnerability has been found in BerriAI LiteLLM up to 1.94.0. This affects the function get_secret of the file secret_managers/main.py of the component Secret Resolution. The manipulation of the argument api_key leads to improper authorization. The attack can be initiated remotely. The exploit has been disclosed to the public and may be used. The vendor was contacted early about this disclosure but did not respond in any way.

  4. CVE-2026-108574 4.3 Moyenne

    BerriAI LiteLLM

    A flaw has been found in BerriAI LiteLLM up to 1.95.0. Affected by this issue is the function ui_view_session_spend_logs of the file litellm/proxy/spend_tracking/spend_management_endpoints.py of the component Spend Tracking. Executing a manipulation of the argument session_id can lead to authorization bypass. It is possible to launch the attack remotely. The exploit has been published and may be used. Upgrading to version 1.96.0 can resolve this issue. This patch is called 722d9ffa4f6c5ae15702ab9ab2c5f6bf1688308b. The affected component should be upgraded.

  5. CVE-2026-93355 8.1 Élevée

    BerriAI litellm

    LiteLLM contains a weak authentication vulnerability that allows an attacker holding a valid JWT from the configured identity provider to authenticate as any existing user by exploiting an email-based fallback lookup in the JWT authentication flow without verifying the email_verified claim. Attackers can present a token with an unverified email address matching a victim's account to inherit the victim's role, including proxy_admin privileges, and permanently overwrite the victim's stored identity binding to retain persistent unauthorized access to administrative endpoints exposing API keys and user management.

  6. CVE-2026-89032 7.7 Élevée

    BerriAI litellm

    BerriAI LiteLLM before 1.101.0-rc.1 contains a tenant isolation bypass vulnerability in the semantic cache layer that allows authenticated users to read other tenants' cached responses by exploiting a metadata key mismatch between _get_semantic_cache_tenant_scope() and _get_metadata_variable_name(). Attackers holding a valid virtual key can submit semantically similar prompts on affected routes such as /v1/responses and /bedrock/* to retrieve cached responses containing other tenants' personally identifiable information, financial data, or source code, and can cause agentic front-ends to auto-execute attacker-supplied tool calls under victim credentials by returning cached function_call or tool_calls payloads to a different principal.

  7. CVE-2026-59823 5.3 Moyenne

    BerriAI litellm

    LiteLLM is a proxy server (AI Gateway) to call LLM APIs in OpenAI (or native) format. Prior to 1.83.9, an authenticated LiteLLM Proxy caller with a valid virtual key can place api_base inside the user_config request body to bypass is_request_body_safe, which blocks top-level api_base and base_url but previously did not inspect or reject user_config. Because user_config constructs the outbound router, the nested destination redirects a server-side request to an internal or external host selected by the caller and can expose endpoints the caller cannot otherwise access. This issue is fixed in version 1.83.9.

  8. CVE-2026-84377 6.5 Moyenne

    BerriAI litellm

    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.

  9. CVE-2026-37004 9.8 Critique

    BerriAI litellm <=1.82.4 is vulnerable to Server-Side Template Injection (SSTI), which allows unauthenticated remote attackers to execute arbitrary OS commands via a crafted dotprompt_content parameter in the /prompts/test endpoint due to use of an unsandboxed jinja2.Environment.

  10. CVE-2026-42208 9.8 Critique Exploitée

    BerriAI litellm · Red Hat Lightspeed Core · Red Hat Ansible Automation Platform 2 · Red Hat OpenShift AI (RHOAI)

    LiteLLM is a proxy server (AI Gateway) to call LLM APIs in OpenAI (or native) format. From version 1.81.16 to before version 1.83.7, a database query used during proxy API key checks mixed the caller-supplied key value into the query text instead of passing it as a separate parameter. An unauthenticated attacker could send a specially crafted Authorization header to any LLM API route (for example POST /chat/completions) and reach this query through the proxy's error-handling path. An attacker could read data from the proxy's database and may be able to modify it, leading to unauthorised access to the proxy and the credentials it manages. This issue has been patched in version 1.83.7.