vLLM in Thailand
The companies, labs and universities running vLLM here. If your team is on this list, you are not doing it alone.
5 organisations
AGICAFET
Bangkok
Organiser of vLLM Bangkok Day. Builds and serves Thai large language models, including the ThaiLLM series.
Serving Thai-language models in production, including NVFP4-quantised checkpoints.
agicafet.comEmbedded LLM
Strategic partner of vLLM Bangkok Day and a core contributor to vLLM. Works on inference optimisation across GPU platforms, parallelism strategy for large MoE and multimodal models, and private on-premise serving.
Upstream vLLM contributions (DFlash, KV-cache compression and offloading, prefix caching, chunked prefill, sleep mode) and production serving on limited GPUs.
embeddedllm.comRed Hat
Enterprise open source. A major contributor to vLLM and to open model serving on Kubernetes.
Upstream vLLM contribution and enterprise inference platforms.
redhat.comTensormesh
Inference infrastructure — making LLM serving faster and cheaper at scale.
Inference optimisation and KV cache systems built around vLLM.
tensormesh.aiInferact
Partner of vLLM Bangkok Day, working on AI inference.