# eagle-hq/EAGLE-4

> Official inference harness for EAGLE-4 speculative decoding, adding a flexible drafter plugin to SGLang, vLLM and TokenSpeed that adapts to batch size, hardware and workload. It speeds up serving of LLMs such as Qwen3.8-27B while preserving the target model's output distribution.

- Magnitude: 5.1 out of 10 — Early signal
- Stars: 137 total · Not enough measurements today
- Star trust: star growth looks organic
- Category: Language models · Language: Python · Created: 2026-10-20 · Last push: 2026-10-20
- GitHub: https://github.com/eagle-hq/EAGLE-4 · Homepage: https://eagle.ai · Page: https://gitnova.dev/en/r/eagle-hq/EAGLE-4

## Useful for

- Speed up Qwen3.8-27B serving by installing the eagle-harness plugin on SGLang or vLLM
- Run the prebuilt Docker image with the patched engine and EAGLE-4 on NVIDIA B300 GPUs
- Verify that the installed engine exposes all hooks the EAGLE-4 plugin needs using eagle-harness-check

## Why it’s here

- There are not enough fresh counter measurements for today's UTC day. The gain is unknown.
- The repository is 1 day old and already has 137 stars. With less than two weeks of history, there's no usual pace to compare the spike against yet.
- Top new repositories this week: #134.

## Star trust

Star growth looks organic.

Star-trust labels are heuristics based on the repository’s behavior, not a check of every stargazer.

## Numbers

- Forks: 2
- Issues and pull requests: 0
- Watchers: 0
- Average over the last week: 142 per day
- Usual pace: too little history (under two weeks)
- Latest release: v0.1.0 (2026-10-20)

## Stars per day, last 2 days (oldest → newest, today is partial)

2026-10-20 … 2026-10-21: 143, None

## Spotted in now

- Top new repositories this week: #134

## More in this category

1. **deepseek-ai/DeepInfer** — 10.0 · Breakout · Language models · Python · Not enough measurements today
   Offline inference framework for running DeepSeek models at large scale: optimized kernels, MoE/Attention parallelism, deterministic sampling and structured outputs. Aimed at teams deploying DeepSeek-V4.1-Flash on NVIDIA SM100 or Huawei…
   Full card: https://gitnova.dev/en/r/deepseek-ai/DeepInfer.md
2. **google-pytorch/torch_tpu** — 7.2 · Early signal · Language models · C++ · Not enough measurements today
   TorchTPU is a PyTorch backend for Google TPUs that lets you run PyTorch models and operations on TPU devices, with custom ATen kernels, an XLA compilation cache, and distributed training support.
   Full card: https://gitnova.dev/en/r/google-pytorch/torch_tpu.md
3. **vllm-project/vllm** — 5.3 · Breakout · Language models · Python · Not enough measurements today
   A high-throughput and memory-efficient inference and serving engine for large language models, using PagedAttention for KV-cache management. Supports 200+ model architectures and an OpenAI-compatible API.
   Full card: https://gitnova.dev/en/r/vllm-project/vllm.md
4. **MiaAI-Lab/DeepSeek-v4.1-Flash-1017-2x-DGX-Sparks** — 5.1 · Early signal · Language models · Shell · Not enough measurements today
   Scripts to serve the quantized DeepSeek-V4.1-Flash-1017 (EXL3, 2.9 bpw) across two NVIDIA DGX Sparks via the TensorFold engine, exposing an OpenAI-compatible API with up to a 1M-token window and DSpark speculative decoding.
   Full card: https://gitnova.dev/en/r/MiaAI-Lab/DeepSeek-v4.1-Flash-1017-2x-DGX-Sparks.md
5. **jatechan/ai-data-extractor** — 5.0 · Early signal · Language models · Python · Not enough measurements today
   Python tool that extracts local chat histories from AI coding assistants (Claude Code, Cursor, Windsurf, Aider, Cline/Roo Code and more) into a unified JSONL format for fine-tuning, analytics, or backup.
   Full card: https://gitnova.dev/en/r/jatechan/ai-data-extractor.md

---

Magnitude (0–10) measures how fast and how unusually interest in a repository is growing right now. It is not a quality score.
Days are UTC. “So far today” is a fact; “expected by the end of the day” is a forecast.
Summaries and use cases are written by an LLM (DeepSeek V4.1 Flash) from the README and may be inaccurate: verify specific claims (benchmarks, speed, hardware) in the repository itself.
Data as of 2026-10-21 00:15 UTC, updated every 30 minutes.