Skip to content
ENTITY PROFILE / KNOWLEDGE GRAPH

DeepInfer

RSS

DeepSeek announced on Oct. 20 that it had open‑sourced the inference framework DeepInfer. DeepInfer currently supports the DeepSeek-V4.1-Flash model. DeepInfer supports NVIDIA SM100 graphics processors. DeepInfer supports Huawei Ascend 950 chips. DeepInfer is based on the vLLM framework DeepInfer is released under the Apache License 2.0 DeepSeek released DeepInfer DeepInfer runs on NVIDIA DeepInfer runs on Huawei Ascend platform DeepInfer supports SWA Bounded Replay DeepInfer uses deepseek-recipe component DeepInfer runs on SM100/A950 card DeepInfer is the inference engine behind V4.1 Flash DeepInfer CUSTOMIZED_WITH SGLang DeepInfer CUSTOMIZED_WITH TRRTLLM DeepInfer SUPPORTS NVIDIA DeepInfer SUPPORTS Huawei Ascend DeepInfer COMPARES_WITH OpenAI· Entity summary generated by Graphiti

RELATIONS
18Relations
CURRENT
18Current
EVENTS
3Events
REPORTS
3Source reports
KNOWLEDGE ATLAS0 entities · 0 relations
No public neighbors yet
Explore all published links of a node; previously discovered relations remain visible.

Relation timeline

BY VALID-FROM DATE
  1. DeepInferRUNS ONNVIDIA

    DeepInfer runs on NVIDIA

    Current1 sourceNVIDIA
  2. DeepInferRUNS ONHuawei Ascend 950 chips

    DeepInfer runs on Huawei Ascend platform

    Current1 sourceHuawei Ascend 950 chips
  3. DeepInferCUSTOMIZED WITHSGLang

    DeepInfer CUSTOMIZED_WITH SGLang

    Current1 sourceSGLang
  4. DeepInferUSESdeepseek-recipe

    DeepInfer uses deepseek-recipe component

    Current1 sourcedeepseek-recipe
  5. DeepInferRELATES TODeepseek V4.1-Flash

    DeepInfer is the inference engine behind V4.1 Flash

    Current1 sourceDeepseek V4.1-Flash
  6. DeepInferSUPPORTSNVIDIA

    DeepInfer SUPPORTS NVIDIA

    Current1 sourceNVIDIA
  7. DeepInferCOMPARES WITHOpenAI

    DeepInfer COMPARES_WITH OpenAI

    Current1 sourceOpenAI
  8. DeepInferSUPPORTSHuawei Ascend 950 chips

    DeepInfer SUPPORTS Huawei Ascend

    Current1 sourceHuawei Ascend 950 chips
  9. DeepInferCUSTOMIZED WITHTRRTLLM

    DeepInfer CUSTOMIZED_WITH TRRTLLM

    Current1 sourceTRRTLLM
  10. DeepseekRELEASEDDeepInfer

    DeepSeek released DeepInfer

    Current2 sourcesDeepseek
  11. DeepInferRUNS ONNVIDIA SM100 graphics processors

    DeepInfer runs on SM100/A950 card

    Current1 sourceNVIDIA SM100 graphics processors
  12. DeepInferSUPPORTSFlashInfer

    DeepInfer supports SWA Bounded Replay

    Current1 sourceFlashInfer
  13. DeepInferRELEASED UNDER LICENSEDeepseek V4.1-Flash-1017

    DeepInfer is released under the Apache License 2.0

    Current1 sourceDeepseek V4.1-Flash-1017
  14. DeepInferBUILDS ONvLLM

    DeepInfer is based on the vLLM framework

    Current3 sourcesvLLM
  15. DeepInferSUPPORTSNVIDIA SM100 graphics processors

    DeepInfer supports NVIDIA SM100 graphics processors.

    Current1 sourceNVIDIA SM100 graphics processors
  16. DeepseekRELEASEDDeepInfer

    DeepSeek announced on Oct. 20 that it had open‑sourced the inference framework DeepInfer.

    Current3 sourcesDeepseek
  17. DeepInferSUPPORTSDeepseek V4.1-Flash

    DeepInfer currently supports the DeepSeek-V4.1-Flash model.

    Current2 sourcesDeepseek V4.1-Flash
  18. DeepInferSUPPORTSHuawei Ascend 950 chips

    DeepInfer supports Huawei Ascend 950 chips.

    Current1 sourceHuawei Ascend 950 chips

Relations are facts Graphiti extracted from reports, not verbatim quotes. An end date means later reporting showed the relation no longer holds.

Events 3

Source reports 3