- 1Releases tracked
- Oct 10Latest · Decision 3.0 (d3, d3-flash, d3-mini, d3-nano, d3-lite and d3-edge)
- 100%Open-weight
- USUnited States
vLLM Semantic Router is an open-source project, maintained by the vLLM Semantic Router Team under the vLLM project's GitHub organization, that builds an open, programmable decision layer for routing requests across models and compute in Mixture-of-Models systems. The router is released under the Apache 2.0 license. Its tracked release, Decision 3.0 (October 2026), is an open-weight multimodal decision model family from 0.6B to 27B parameters, comprising d3, d3-flash, d3-mini, d3-nano, d3-lite and d3-edge. The models read text, images and video and return probabilities rather than generating text.
vLLM Semantic Router
1 release latest Oct 10, 2026
-
Decision 3.0 (d3, d3-flash, d3-mini, d3-nano, d3-lite and d3-edge)
Open Apache-2.0 multimodal decision family (0.6B–27B) reading text, images and video; System One probabilities, no text generation.
vLLM Semantic Router FAQ
- What is vLLM Semantic Router's latest AI model?
- vLLM Semantic Router's most recent release in the Startrise tracker is Decision 3.0 (d3, d3-flash, d3-mini, d3-nano, d3-lite and d3-edge), released Oct 10, 2026.
- How many AI models has vLLM Semantic Router released?
- Startrise tracks 1 notable vLLM Semantic Router release since January 2025, from Decision 3.0 (d3, d3-flash, d3-mini, d3-nano, d3-lite and d3-edge) (Oct 10, 2026) to Decision 3.0 (d3, d3-flash, d3-mini, d3-nano, d3-lite and d3-edge).
- Does vLLM Semantic Router release open-weight models?
- Yes. 1 of 1 tracked vLLM Semantic Router releases (100%) ship with downloadable weights; the most recent is Decision 3.0 (d3, d3-flash, d3-mini, d3-nano, d3-lite and d3-edge).
- Where is vLLM Semantic Router based?
- vLLM Semantic Router is based in United States.