Berriai
Products
1- Litellm51 CVEspypi
Recent CVEs
51| CVE | Vendor / Product | Sev | Risk | CVSS | EPSS | KEV | Published | Description |
|---|---|---|---|---|---|---|---|---|
| CVE-2026-42208 | Cri | 0.79 | 9.8 | 0.89 | KEV | May 8, 2026 | 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… | |
| CVE-2026-42271 | Hig | 0.69 | 8.8 | 0.91 | KEV | May 8, 2026 | LiteLLM is a proxy server (AI Gateway) to call LLM APIs in OpenAI (or native) format. From version 1.74.2 to before version 1.83.7, two endpoints used to preview an MCP server before saving it — POST /mcp-rest/test/connection and POST /mcp-rest/test/tools/list — accepted a… | |
| CVE-2026-33634 | Hig | 0.67 | 8.8 | 0.62 | KEV | Mar 23, 2026 | Trivy is a security scanner. On March 19, 2026, a threat actor used compromised credentials to publish a malicious Trivy v0.69.4 release, force-push 76 of 77 version tags in `aquasecurity/trivy-action` to credential-stealing malware, and replace all 7 tags in… | |
| CVE-2026-123061 | cri | 0.64 | 9.8 | — | Oct 27, 2026 | litellm: litellm: unauthenticated arbitrary file read leading to master key leakage and remote code execution | ||
| CVE-2026-30623 | Cri | 0.64 | 9.8 | 0.06 | Jul 15, 2026 | LiteLLM 1.18.10 contains a remote code execution vulnerability in its MCP server creation functionality. The application allows users to add MCP servers via a JSON configuration specifying arbitrary command and args values. LiteLLM executes these values on the host without… | ||
| CVE-2024-4264 | Cri | 0.64 | 9.8 | 0.01 | May 18, 2024 | A remote code execution (RCE) vulnerability exists in the berriai/litellm project due to improper control of the generation of code when using the `eval` function unsafely in the `litellm.get_secret()` method. Specifically, when the server utilizes Google KMS, untrusted data is… | ||
| CVE-2026-40217 | Hig | 0.59 | 8.8 | 0.78 | KEV | Apr 10, 2026 | LiteLLM through 2026-04-08 allows remote attackers to execute arbitrary code via bytecode rewriting at the /guardrails/test_custom_code URI. | |
| CVE-2026-59822 | Hig | 0.58 | 8.2 | 0.01 | KEV | Jul 8, 2026 | LiteLLM is a proxy server (AI Gateway) to call LLM APIs in OpenAI (or native) format. Prior to 1.84.0, LiteLLM's MCP Streamable HTTP endpoint allowed an unauthenticated attacker to use a fabricated Authorization header to trigger an OAuth2 passthrough fallback path that replaced… | |
| CVE-2026-114612 | Cri | 0.57 | 9.9 | 0.00 | Oct 22, 2026 | 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… | ||
| CVE-2026-37004 | Cri | 0.57 | 9.8 | 0.01 | Aug 27, 2026 | 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. | ||
| CVE-2026-49468 | Cri | 0.57 | 9.8 | 0.03 | Jun 22, 2026 | LiteLLM is a proxy server (AI Gateway) to call LLM APIs in OpenAI (or native) format. Prior to 1.84.0, a Host-header parsing flaw in the LiteLLM proxy could, under specific conditions, allow unauthenticated access to protected management routes. The auth layer derived the… | ||
| CVE-2026-42203 | Hig | 0.57 | 8.8 | 0.00 | May 8, 2026 | LiteLLM is a proxy server (AI Gateway) to call LLM APIs in OpenAI (or native) format. From version 1.80.5 to before version 1.83.7, the POST /prompts/test endpoint accepted user-supplied prompt templates and rendered them without sandboxing. A crafted template could run… | ||
| CVE-2024-5751 | Cri | 0.57 | 9.8 | 0.01 | Jun 27, 2024 | BerriAI/litellm version v1.35.8 contains a vulnerability where an attacker can achieve remote code execution. The vulnerability exists in the `add_deployment` function, which decodes and decrypts environment variables from base64 and assigns them to `os.environ`. An attacker can… | ||
| CVE-2024-2952 | Cri | 0.57 | 9.8 | 0.01 | Apr 10, 2024 | BerriAI/litellm is vulnerable to Server-Side Template Injection (SSTI) via the `/completions` endpoint. The vulnerability arises from the `hf_chat_template` method processing the `chat_template` parameter from the `tokenizer_config.json` file through the Jinja template engine… | ||
| CVE-2026-93355 | Hig | 0.53 | 8.1 | 0.00 | Sep 28, 2026 | 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… | ||
| CVE-2026-35030 | Cri | 0.52 | 9.1 | 0.01 | Apr 6, 2026 | LiteLLM is a proxy server (AI Gateway) to call LLM APIs in OpenAI (or native) format. Prior to 1.83.0, when JWT authentication is enabled (enable_jwt_auth: true), the OIDC userinfo cache uses token[:20] as the cache key. JWT headers produced by the same signing algorithm… | ||
| CVE-2026-35029 | Hig | 0.52 | 8.8 | 0.39 | Apr 6, 2026 | LiteLLM is a proxy server (AI Gateway) to call LLM APIs in OpenAI (or native) format. Prior to 1.83.0, the /config/update endpoint does not enforce admin role authorization. A user who is already authenticated into the platform can then use this endpoint to modify proxy… | ||
| CVE-2026-47102 | Hig | 0.50 | 8.8 | 0.01 | May 21, 2026 | LiteLLM prior to 1.83.10 allows a user to modify their own user_role via the /user/update endpoint. While the endpoint correctly restricts users to updating only their own account, it does not restrict which fields may be changed. A user who can reach this endpoint can set their… | ||
| CVE-2026-47101 | Hig | 0.50 | 8.8 | 0.01 | May 21, 2026 | LiteLLM prior to 1.83.14 allows an authenticated internal_user to create API keys with access to routes that their role does not permit. When generating a key, the allowed_routes field is stored without verifying that the specified routes fall within the user's own permissions.… | ||
| CVE-2024-6825 | Hig | 0.50 | 8.8 | 0.02 | Mar 20, 2025 | BerriAI/litellm version 1.40.12 contains a vulnerability that allows remote code execution. The issue exists in the handling of the 'post_call_rules' configuration, where a callback function can be added. The provided value is split at the final '.' mark, with the last part… |
- risk 0.79cvss 9.8epss 0.89
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…
- risk 0.69cvss 8.8epss 0.91
LiteLLM is a proxy server (AI Gateway) to call LLM APIs in OpenAI (or native) format. From version 1.74.2 to before version 1.83.7, two endpoints used to preview an MCP server before saving it — POST /mcp-rest/test/connection and POST /mcp-rest/test/tools/list — accepted a…
- risk 0.67cvss 8.8epss 0.62
Trivy is a security scanner. On March 19, 2026, a threat actor used compromised credentials to publish a malicious Trivy v0.69.4 release, force-push 76 of 77 version tags in `aquasecurity/trivy-action` to credential-stealing malware, and replace all 7 tags in…
- risk 0.64cvss 9.8epss —
litellm: litellm: unauthenticated arbitrary file read leading to master key leakage and remote code execution
- risk 0.64cvss 9.8epss 0.06
LiteLLM 1.18.10 contains a remote code execution vulnerability in its MCP server creation functionality. The application allows users to add MCP servers via a JSON configuration specifying arbitrary command and args values. LiteLLM executes these values on the host without…
- risk 0.64cvss 9.8epss 0.01
A remote code execution (RCE) vulnerability exists in the berriai/litellm project due to improper control of the generation of code when using the `eval` function unsafely in the `litellm.get_secret()` method. Specifically, when the server utilizes Google KMS, untrusted data is…
- risk 0.59cvss 8.8epss 0.78
LiteLLM through 2026-04-08 allows remote attackers to execute arbitrary code via bytecode rewriting at the /guardrails/test_custom_code URI.
- risk 0.58cvss 8.2epss 0.01
LiteLLM is a proxy server (AI Gateway) to call LLM APIs in OpenAI (or native) format. Prior to 1.84.0, LiteLLM's MCP Streamable HTTP endpoint allowed an unauthenticated attacker to use a fabricated Authorization header to trigger an OAuth2 passthrough fallback path that replaced…
- risk 0.57cvss 9.9epss 0.00
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…
- risk 0.57cvss 9.8epss 0.01
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.
- risk 0.57cvss 9.8epss 0.03
LiteLLM is a proxy server (AI Gateway) to call LLM APIs in OpenAI (or native) format. Prior to 1.84.0, a Host-header parsing flaw in the LiteLLM proxy could, under specific conditions, allow unauthenticated access to protected management routes. The auth layer derived the…
- risk 0.57cvss 8.8epss 0.00
LiteLLM is a proxy server (AI Gateway) to call LLM APIs in OpenAI (or native) format. From version 1.80.5 to before version 1.83.7, the POST /prompts/test endpoint accepted user-supplied prompt templates and rendered them without sandboxing. A crafted template could run…
- risk 0.57cvss 9.8epss 0.01
BerriAI/litellm version v1.35.8 contains a vulnerability where an attacker can achieve remote code execution. The vulnerability exists in the `add_deployment` function, which decodes and decrypts environment variables from base64 and assigns them to `os.environ`. An attacker can…
- risk 0.57cvss 9.8epss 0.01
BerriAI/litellm is vulnerable to Server-Side Template Injection (SSTI) via the `/completions` endpoint. The vulnerability arises from the `hf_chat_template` method processing the `chat_template` parameter from the `tokenizer_config.json` file through the Jinja template engine…
- risk 0.53cvss 8.1epss 0.00
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…
- risk 0.52cvss 9.1epss 0.01
LiteLLM is a proxy server (AI Gateway) to call LLM APIs in OpenAI (or native) format. Prior to 1.83.0, when JWT authentication is enabled (enable_jwt_auth: true), the OIDC userinfo cache uses token[:20] as the cache key. JWT headers produced by the same signing algorithm…
- risk 0.52cvss 8.8epss 0.39
LiteLLM is a proxy server (AI Gateway) to call LLM APIs in OpenAI (or native) format. Prior to 1.83.0, the /config/update endpoint does not enforce admin role authorization. A user who is already authenticated into the platform can then use this endpoint to modify proxy…
- risk 0.50cvss 8.8epss 0.01
LiteLLM prior to 1.83.10 allows a user to modify their own user_role via the /user/update endpoint. While the endpoint correctly restricts users to updating only their own account, it does not restrict which fields may be changed. A user who can reach this endpoint can set their…
- risk 0.50cvss 8.8epss 0.01
LiteLLM prior to 1.83.14 allows an authenticated internal_user to create API keys with access to routes that their role does not permit. When generating a key, the allowed_routes field is stored without verifying that the specified routes fall within the user's own permissions.…
- risk 0.50cvss 8.8epss 0.02
BerriAI/litellm version 1.40.12 contains a vulnerability that allows remote code execution. The issue exists in the handling of the 'post_call_rules' configuration, where a callback function can be added. The provided value is split at the final '.' mark, with the last part…