# SpecFit: Not All Draft Tokens Cost the Same in Offloaded MoE Inference

arXiv: https://arxiv.org/abs/2610.15417 · HTML: https://mlsanity.com/paper/2610.15417/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Efficient MoE training and inference systems, one of 12 directions in Mixture-of-experts.
- Builds on 21 of its 29 references that are in the map.
- First posted on arXiv 11 Oct 2026.

## Placed in

- Mixture-of-experts / [Efficient MoE training and inference systems](https://mlsanity.com/moe/efficient-moe-training-and-inference-systems/)
- Efficient LLM inference and serving / [Speculative decoding](https://mlsanity.com/inference/speculative-decoding/)

## Builds on (most cited first, 21 of 21)

- [Training Verifiers to Solve Math Word Problems](https://mlsanity.com/paper/2110.14168/) (2021)
- [Outrageously Large Neural Networks: The Sparsely-Gated Mixture-of-Experts Layer](https://mlsanity.com/paper/1701.06538/) (2017)
- [Switch Transformers: Scaling to Trillion Parameter Models with Simple and Efficient Sparsity](https://mlsanity.com/paper/2101.03961/) (2021)
- [Fast Inference from Transformers via Speculative Decoding](https://mlsanity.com/paper/2211.17192/) (2022)
- [Accelerating Large Language Model Decoding with Speculative Sampling](https://mlsanity.com/paper/2302.01318/) (2023)
- [Abstractive Text Summarization Using Sequence-to-Sequence RNNs and Beyond](https://mlsanity.com/paper/1602.06023/) (2016)
- [EAGLE-2: Faster Inference of Language Models with Dynamic Draft Trees](https://mlsanity.com/paper/2406.16858/) (2024)
- [EAGLE-3: Scaling up Inference Acceleration of Large Language Models via Training-Time Test](https://mlsanity.com/paper/2503.01840/) (2025)
- [Unlocking Efficiency in Large Language Model Inference: A Comprehensive Survey of Speculative Decoding](https://mlsanity.com/paper/2401.07851/) (2024)
- [Fast Inference of Mixture-of-Experts Language Models with Offloading](https://mlsanity.com/paper/2312.17238/) (2023)
- [Fiddler: CPU-GPU Orchestration for Fast Inference of Mixture-of-Experts Models](https://mlsanity.com/paper/2402.07033/) (2024)
- [OPT-Tree: Speculative Decoding with Adaptive Draft Tree Structure](https://mlsanity.com/paper/2406.17276/) (2024)
- [MoE-Infinity: Efficient MoE Inference on Personal Machines with Sparsity-Aware Expert Cache](https://mlsanity.com/paper/2401.14361/) (2024)
- [SpecExec: Massively Parallel Speculative Decoding for Interactive LLM Inference on Consumer Devices](https://mlsanity.com/paper/2406.02532/) (2024)
- [ProMoE: Fast MoE-based LLM Serving using Proactive Caching](https://mlsanity.com/paper/2410.22134/) (2024)
- [SP-MoE: Speculative Decoding and Prefetching for Accelerating MoE-based Model Inference](https://mlsanity.com/paper/2510.10302/) (2025)
- [MoE-Spec: Expert Budgeting for Efficient Speculative Decoding](https://mlsanity.com/paper/2602.16052/) (2026)
- [ETS: Energy-Guided Test-Time Scaling for Training-Free RL Alignment](https://mlsanity.com/paper/2601.21484/) (2026)
- [Pre-Attention Expert Prediction and Prefetching for Mixture-of-Experts Large Language Models](https://mlsanity.com/paper/2511.10676/) (2025)
- [Turning Off-Policy Tokens On-Policy: A Plug-in Approach for Improving LLM Alignment](https://mlsanity.com/paper/2607.04728/) (2026)
- [DraftExpert: Expansion-Aware Self-Speculative Decoding for End-Device MoE Inference](https://mlsanity.com/paper/2607.24434/) (2026)

## Abstract (arXiv)

Mixture-of-experts (MoE) models activate only a few experts per token, but when their expert weights are offloaded to host memory, host-to-device (H2D) expert transfers enter the critical path of decoding. Speculative decoding verifies several draft tokens in one forward pass, yet in an MoE model each verified token activates its own experts, so a larger draft tree transfers more experts. Because of this H2D cost, speculative decoding commits several tokens per verification cycle yet is sometimes even slower than autoregressive decoding. Based on this observation, we propose SpecFit, which chooses the verified draft subtree to maximize predicted throughput rather than acceptance length. SpecFit estimates the expected number of committed tokens from draft path probabilities and predicts cycle time from the node count and the per-layer union of the experts the subtree activates; this union is a monotone submodular coverage function, so each node pays only for the experts it adds. An exact search over ancestor-closed subtrees returns the subtree with the highest predicted throughput; run after each draft level, the same search also decides when to stop drafting, and the unmodified target verifies the result. On CNN/DailyMail summarization and a GSM8K subset with Qwen3-30B-A3B, SpecFit decodes 21.6% and 7.0% faster than EAGLE-3 with the same expert cache and 1.97x and 1.72x as fast as vanilla EAGLE-3. Code is available at https://github.com/Chlience/SpecFit.