Search Results (18 CVEs found)

CVE Vendors Products Updated CVSS v3.1
CVE-2026-85695 2 Lm-sys, Lmsys 2 Fastchat, Fastchat 2026-10-01 9.4 Critical
FastChat contains an authentication bypass vulnerability in the /register_worker endpoint that allows unauthenticated attackers to register arbitrary worker addresses and perform server-side request forgery. Attackers can register malicious workers under victim model names to intercept user prompts, images, and responses, or probe internal network ports across the worker mesh.
CVE-2026-102634 1 Lmsys 1 Sglang 2026-09-29 7.5 High
SGLang through 0.5.20 in prefill/decode disaggregation mode fails to validate duplicate bootstrap_room fields in /generate requests with Mooncake KV transfer backend. Unauthenticated attackers can send concurrent requests with identical bootstrap_room values to crash scheduler processes or hang other users' requests until transfer timeout.
CVE-2026-93688 2 Lmsys, Sgl-project 2 Sglang, Sglang 2026-09-21 7.5 High
SGLang through 0.5.19 in prefill/decode disaggregation mode with Mooncake KV transfer backend fails to validate bootstrap_room values, allowing unbounded transfer state allocation. Unauthenticated attackers can reach the decode engine's POST /generate endpoint and submit arbitrary bootstrap_room values to exhaust prefill process memory until out-of-memory termination.
CVE-2026-93838 2 Lmsys, Sgl-project 2 Sglang, Sglang 2026-09-20 5.9 Medium
SGLang versions through 0.5.20 contain an unbounded memory allocation vulnerability in handle_staging_req() that fails to validate chunk_idx from ZMQ STAGING_REQ frames in prefill/decode disaggregation deployments. Attackers with access to the decode engine's internal ZMQ rank port can send a frame with an extremely large chunk_idx value, causing the scheduler to allocate memory until the system runs out and terminates the process.
CVE-2026-15978 2 Lmsys, Sglang 2 Sglang, Sglang 2026-08-04 7.5 High
SGLang contains a model weight exfiltration vulnerability when no API keys are configured, as SGLang will expose two endpoints that allow a remote attacker to trigger distributed weight broadcasting using NCCL and then triggering data transfer, attackers can exfiltrate all model weights.
CVE-2026-15971 2 Lmsys, Sglang 2 Sglang, Sglang 2026-08-02 9.8 Critical
SGLang contains an RCE vulnerability when the optional dumper subsystem is enabled, allowing for a sandbox escape when DUMPER_SERVER_PORT is set, enabling code execution on inference requests.
CVE-2026-15976 2 Lmsys, Sglang 2 Sglang, Sglang 2026-07-31 9.8 Critical
SGLang contains a RCE vulnerability when attempting to load model weights from a HuggingFace repository, specifically within the /update_weights_from_disk, where torch.load(..., weights_only=False) fallback enables pickle deserialization of .bin files.
CVE-2026-15974 2 Lmsys, Sglang 2 Sglang, Sglang 2026-07-31 6.5 Medium
SGLang contains an SSRF and local file read in the multimodal generation endpoint /v1/chat/completions due to unsanitized image_url, allowing access to internal metadata, secrets, and services.
CVE-2026-15977 2 Lmsys, Sglang 2 Sglang, Sglang 2026-07-31 7.5 High
SGLang contains a credential leakage vulnerability in the /server_info endpoint, which will return API keys and SSL keyfile information when only the --admin-api-key is configured.
CVE-2026-15969 2 Lmsys, Sglang 2 Sglang, Sglang 2026-07-31 9.8 Critical
SGLang contains an unauthenticated RCE in /load_lora_adapter_from_tensors via bypass of SafeUnpickler’s incomplete denylist, allowing arbitrary command execution through crafted base64-encoded pickle payloads.
CVE-2026-14890 2 Lmsys, Sglang 2 Sglang, Sglang 2026-07-16 9.1 Critical
SGLang uses an expert-parallel backup subsystem that exposes a ZeroMQ PULL socket on a routable network interface that does not contain authentication or deserialization safeguards, allowing an attacker to provide a malicious pickle file that results in unauthenticated remote code execution when the feature is enabled and the service is reachable over the network.
CVE-2026-5760 2 Lmsys, Sglang 2 Sglang, Sglang 2026-06-03 9.8 Critical
SGLang's reranking endpoint (/v1/rerank) achieves Remote Code Execution (RCE) when a model file containing a malcious tokenizer.chat_template is loaded, as the Jinja2 chat templates are rendered using an unsandboxed jinja2.Environment().
CVE-2026-7301 2 Lmsys, Sglang 2 Sglang, Sglang 2026-05-19 9.8 Critical
SGLangs multimodal generation runtime scheduler's ROUTER socket binds to 0.0.0.0 by default and contains a sink that calls pickle.loads() on incoming messages, enabling RCE when exposed to the internet.
CVE-2026-7302 2 Lmsys, Sglang 2 Sglang, Sglang 2026-05-19 9.1 Critical
SGLangs multimodal generation runtime is vulnerable to an unauthenticated path traversal vulnerability, allowing an attacker to write arbitrary files anywhere the server process has write access, by including ../ sequences in the upload filename when sent to specific endpoints.
CVE-2026-7304 2 Lmsys, Sglang 2 Sglang, Sglang 2026-05-19 9.8 Critical
SGLangs multimodal generation runtime is vulnerable to unauthenticated remote code execution when the --enable-custom-logit-processor option is enabled, as Python objects loaded via dill.loads() will be deserialized without validation.
CVE-2026-3989 2 Lmsys, Sglang 2 Sglang, Sglang 2026-04-07 7.8 High
SGLangs `replay_request_dump.py` contains an insecure pickle.load() without validation and proper deserialization. An attacker can take advantage of this by providing a malicious .pkl file, which will execute the attackers code on the device running the script.
CVE-2026-3060 2 Lmsys, Sglang 2 Sglang, Sglang 2026-04-07 9.8 Critical
SGLang' encoder parallel disaggregation system is vulnerable to unauthenticated remote code execution through the disaggregation module, which deserializes untrusted data using pickle.loads() without authentication.
CVE-2026-3059 2 Lmsys, Sglang 2 Sglang, Sglang 2026-04-07 9.8 Critical
SGLang's multimodal generation module is vulnerable to unauthenticated remote code execution through the ZMQ broker, which deserializes untrusted data using pickle.loads() without authentication.