CVE-2024-12704
Denial of Service (DoS) in run-llama/llama_index
A vulnerability in the LangChainLLM class of the run-llama/llama_index repository, version v0.12.5, allows for a Denial of Service (DoS) attack. The stream_complete method executes the llm using a thread and retrieves the result via the get_response_gen method of the StreamingGeneratorCallbackHandler class. If the thread terminates abnormally before the _llm.predict is executed, there is no exception handling for this case, leading to an infinite loop in the get_response_gen function. This can be triggered by providing an input of an incorrect type, causing the thread to terminate and the process to continue running indefinitely.
Attack profile
How easy it is to exploit
- Attack vector
- Network
- Complexity
- Low
- Privileges needed
- None
- User interaction
- None
What an attacker can affect
- Confidentiality
- None
- Integrity
- None
- Availability
- High
Exploit likelihood
What each source reports
CVE List Score used
Denial of Service (DoS) in run-llama/llama_index
A vulnerability in the LangChainLLM class of the run-llama/llama_index repository, version v0.12.5, allows for a Denial of Service (DoS) attack. The stream_complete method executes the llm using a thread and retrieves the result via the get_response_gen method of the StreamingGeneratorCallbackHandler class. If the thread terminates abnormally before the _llm.predict is executed, there is no exception handling for this case, leading to an infinite loop in the get_response_gen function. This can be triggered by providing an input of an incorrect type, causing the thread to terminate and the process to continue running indefinitely.
- Record
CVE-2024-12704- CVSS 3.0
- 7.5 High
CVSS:3.0/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H - Weakness
- CWE-835
- Updated
- 2026-11-02
EUVDEUVD-2025-7006
A vulnerability in the LangChainLLM class of the run-llama/llama_index repository, version v0.12.5, allows for a Denial of Service (DoS) attack. The stream_complete method executes the llm using a thread and retrieves the result via the get_response_gen method of the StreamingGeneratorCallbackHandler class. If the thread terminates abnormally before the _llm.predict is executed, there is no exception handling for this case, leading to an infinite loop in the get_response_gen function. This can be triggered by providing an input of an incorrect type, causing the thread to terminate and the process to continue running indefinitely.
- Record
EUVD-2025-7006- CVSS 3.0
- 7.5 High
CVSS:3.0/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H - Updated
- 2026-11-02
GitHub AdvisoriesGHSA-j3wr-m6xh-64hg
LlamaIndex Improper Handling of Exceptional Conditions vulnerability
A vulnerability in the LangChainLLM class of the run-llama/llama_index repository, version v0.12.5, allows for a Denial of Service (DoS) attack. The stream_complete method executes the llm using a thread and retrieves the result via the get_response_gen method of the StreamingGeneratorCallbackHandler class. If the thread terminates abnormally before the _llm.predict is executed, there is no exception handling for this case, leading to an infinite loop in the get_response_gen function. This can be triggered by providing an input of an incorrect type, causing the thread to terminate and the process to continue running indefinitely.
- Record
GHSA-j3wr-m6xh-64hg- CVSS 3.0
- 7.5 High
CVSS:3.0/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H - Weakness
- CWE-755, CWE-835
- Updated
- 2026-11-02
NVD
A vulnerability in the LangChainLLM class of the run-llama/llama_index repository, version v0.12.5, allows for a Denial of Service (DoS) attack. The stream_complete method executes the llm using a thread and retrieves the result via the get_response_gen method of the StreamingGeneratorCallbackHandler class. If the thread terminates abnormally before the _llm.predict is executed, there is no exception handling for this case, leading to an infinite loop in the get_response_gen function. This can be triggered by providing an input of an incorrect type, causing the thread to terminate and the process to continue running indefinitely.
- Record
CVE-2024-12704- CVSS 3.0
- 7.5 High
CVSS:3.0/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H - Weakness
- CWE-835
- Updated
- 2026-11-02
OSV
Denial of Service (DoS) in run-llama/llama_index
A vulnerability in the LangChainLLM class of the run-llama/llama_index repository, version v0.12.5, allows for a Denial of Service (DoS) attack. The stream_complete method executes the llm using a thread and retrieves the result via the get_response_gen method of the StreamingGeneratorCallbackHandler class. If the thread terminates abnormally before the _llm.predict is executed, there is no exception handling for this case, leading to an infinite loop in the get_response_gen function. This can be triggered by providing an input of an incorrect type, causing the thread to terminate and the process to continue running indefinitely.
- Record
CVE-2024-12704- CVSS 3.0
- 7.5 High
CVSS:3.0/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H - Weakness
- CWE-835
- Updated
- 2026-11-02
OSVGHSA-j3wr-m6xh-64hg
LlamaIndex Improper Handling of Exceptional Conditions vulnerability
A vulnerability in the LangChainLLM class of the run-llama/llama_index repository, version v0.12.5, allows for a Denial of Service (DoS) attack. The stream_complete method executes the llm using a thread and retrieves the result via the get_response_gen method of the StreamingGeneratorCallbackHandler class. If the thread terminates abnormally before the _llm.predict is executed, there is no exception handling for this case, leading to an infinite loop in the get_response_gen function. This can be triggered by providing an input of an incorrect type, causing the thread to terminate and the process to continue running indefinitely.
- Record
GHSA-j3wr-m6xh-64hg- CVSS 3.0
- 7.5 High
CVSS:3.0/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H - Weakness
- CWE-755, CWE-835
- Updated
- 2026-11-02
OSVPYSEC-2025-244
A vulnerability in the LangChainLLM class of the run-llama/llama_index repository, version v0.12.5, allows for a Denial of Service (DoS) attack. The stream_complete method executes the llm using a thread and retrieves the result via the get_response_gen method of the StreamingGeneratorCallbackHandler class. If the thread terminates abnormally before the _llm.predict is executed, there is no exception handling for this case, leading to an infinite loop in the get_response_gen function. This can be triggered by providing an input of an incorrect type, causing the thread to terminate and the process to continue running indefinitely.
- Record
PYSEC-2025-244- CVSS 3.0
- 7.5 High
CVSS:3.0/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H - Updated
- 2026-11-02
OSVPYSEC-2026-1563
LlamaIndex Improper Handling of Exceptional Conditions vulnerability
A vulnerability in the LangChainLLM class of the run-llama/llama_index repository, version v0.12.5, allows for a Denial of Service (DoS) attack. The stream_complete method executes the llm using a thread and retrieves the result via the get_response_gen method of the StreamingGeneratorCallbackHandler class. If the thread terminates abnormally before the _llm.predict is executed, there is no exception handling for this case, leading to an infinite loop in the get_response_gen function. This can be triggered by providing an input of an incorrect type, causing the thread to terminate and the process to continue running indefinitely.
- Record
PYSEC-2026-1563- CVSS 3.0
- 7.5 High
CVSS:3.0/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H - Updated
- 2026-11-02
Affected products
Affected packages
| Package | Affected | Fixed |
|---|---|---|
llama-index-corePyPI |
before 0.12.6 | 0.12.6 |
llama-index-corePyPI |
before 0.12.6, plus listed versions 0.10.00.10.10.10.100.10.110.10.11.post1 Show 45 moreShow less0.10.120.10.130.10.140.10.14.post10.10.150.10.160.10.16.post10.10.170.10.180.10.18.post10.10.190.10.20.10.200.10.20.post10.10.20.post20.10.20.post30.10.210.10.21.post10.10.220.10.230.10.23.post10.10.240.10.24.post10.10.24a10.10.250.10.25.post10.10.25.post20.10.25.post30.10.25a10.10.260.10.270.10.280.10.290.10.30.10.300.10.310.10.320.10.330.10.340.10.350.10.35.post10.10.360.10.370.10.37.post10.10.38 |
0.12.6 |
llama-indexPyPI |
through 0.12.5, plus listed versions 0.10.00.10.10.10.100.10.110.10.12 Show 45 moreShow less0.10.130.10.13.post10.10.140.10.150.10.160.10.170.10.180.10.190.10.200.10.220.10.230.10.240.10.250.10.260.10.270.10.280.10.290.10.30.10.300.10.310.10.320.10.330.10.340.10.350.10.360.10.370.10.380.10.390.10.40.10.400.10.410.10.420.10.430.10.440.10.450.10.45.post10.10.460.10.470.10.480.10.48.post10.10.490.10.50.10.500.10.510.10.52 |
None known |
History
| Recorded | Changed | Sources |
|---|---|---|
| first indexed | CVE List, EUVD, GitHub Advisories, NVD, OSV |
Dates are when Every CVE recorded each change. Upstream dates may differ.
In short
CVE-2024-12704 is a high vulnerability (CVSS 7.5) in Llamaindex, published March 20, 2025. No source lists public exploit code. EPSS puts the chance of exploitation in the next 30 days at 0.83%, the 57th percentile. Fixed in llama-index-core 0.12.6.
References
huntr.com/bounties/a0b638fd-21c6-4ba7-b381-6ab98472a02a
github.com/run-llama/llama_index/commit/d1ecfb77578d089cbe66728f18f635c09aa32a05
nvd.nist.gov/vuln/detail/CVE-2024-12704
pypi.org/project/llama-index-core
Left out here: 489 versions. The JSON record has them.