compare
TensorRT-LLM vs Xinference
The same facts for both, read from GitHub every night, and the relation a person reviewed.
| Fact | TensorRT-LLM | Xinference |
|---|---|---|
| Language | Python | Python |
| Licence | Other | Apache-2.0 |
| Stars | 15k | 9.6k |
| Latest | v1.2.1 | v3.5.0 |
| Last push | 2026-10-06 | 2026-10-06 |
| Release cadence | too few releases | about 15 days between releases |
| Active contributors | 125+ commit authors on the default branch in the last 90 days | 38 commit authors on the default branch in the last 90 days |
| Flags | none | none |
How they relate
Both replace ChatGPT. Alternatives to ChatGPT →
PartialTensorRT-LLMServes open-weight models on NVIDIA GPUs behind an OpenAI-compatible API.
PartialXinferenceServes open-weight language, embedding and speech models behind an OpenAI-compatible API.
TensorRT-LLM
- v1.3.0rc292026-09-29Model Supportpre-release
- v1.3.0rc282026-09-23Known Issuespre-release
- v1.3.0rc272026-09-18Known Issuespre-release
- v1.3.0rc262026-09-09Known Issuespre-release
- v1.3.0rc252026-08-31Annoucementspre-release
Xinference
TensorRT-LLM
✓ signed The latest release, v1.2.1, carries a signature GitHub verified.
Loading the security report
Xinference
✓ signed The latest release, v3.5.0, carries a signature GitHub verified.
Loading the security report