# Efficient LLM inference and serving

3,186 papers in 12 research directions. Data through 7 Oct 2026.
HTML: https://mlsanity.com/inference/

## Directions (largest first)

- [KV cache optimization](https://mlsanity.com/inference/kv-cache-optimization/index.md): 708 papers, 7 new this week
- [Speculative decoding](https://mlsanity.com/inference/speculative-decoding/index.md): 522 papers, 9 new this week
- [Visual token pruning in multimodal models](https://mlsanity.com/inference/visual-token-pruning-in-multimodal-models/index.md): 520 papers, 4 new this week
- [Serving systems and request scheduling](https://mlsanity.com/inference/serving-systems-and-request-scheduling/index.md): 402 papers, 0 new this week
- [Early exit and edge inference](https://mlsanity.com/inference/early-exit-and-edge-inference/index.md): 215 papers, 0 new this week
- [Efficient reasoning and agent inference](https://mlsanity.com/inference/efficient-reasoning-and-agent-inference/index.md): 172 papers, 5 new this week
- [Diffusion LLM and MoE inference](https://mlsanity.com/inference/diffusion-llm-and-moe-inference/index.md): 156 papers, 2 new this week
- [Prompt and context compression](https://mlsanity.com/inference/prompt-and-context-compression/index.md): 132 papers, 0 new this week
- [KV cache quantization](https://mlsanity.com/inference/kv-cache-quantization/index.md): 122 papers, 4 new this week
- [Token pruning and merging for ViTs](https://mlsanity.com/inference/token-pruning-and-merging-for-vits/index.md): 108 papers, 2 new this week
- [Fast autoregressive visual generation](https://mlsanity.com/inference/fast-autoregressive-visual-generation/index.md): 97 papers, 0 new this week
- [Audio and omni-modal token compression](https://mlsanity.com/inference/audio-and-omni-modal-token-compression/index.md): 32 papers, 0 new this week