# HDLFORGE: A Two-Stage Multi-Agent Framework for Efficient Verilog Code Generation with Adaptive Model Escalation

arXiv: https://arxiv.org/abs/2603.04646 · HTML: https://mlsanity.com/paper/2603.04646/

## Key facts

- Cited by 4 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 4 Mar 2026.

## Built on by (most cited first, 4 of 4)

- [COEVO: Co-Evolutionary Framework for Joint Functional Correctness and PPA Optimization in LLM-Based RTL Generation](https://mlsanity.com/paper/2604.15001/) (2026)
- [TimingLLM: A Two-Stage Retrieval-Augmented Framework for Pre-Synthesis Timing Prediction from Verilog](https://mlsanity.com/paper/2604.23602/) (2026)
- [Multi-Level Modeling of Large Language Model Inference Latency and Energy via Hybrid Analytical--Machine-Learning Predictors](https://mlsanity.com/paper/2608.06723/) (2026)
- [What Can Component-Replacement Evidence Establish? A Critical Scoping Review of Local Decisions in LLM Agents](https://mlsanity.com/paper/2609.39989/) (2026)

## Abstract (arXiv)

We present HDLFORGE, a two-stage multi-agent framework for automated Verilog generation that optimizes the trade-off between generation speed and accuracy. The system uses a compact coder with a medium-sized LLM by default (Stage A) and escalates to a stronger coder with an ultra-large LLM (Stage B) only when needed, guided by a calibrated score from inexpensive diagnostics including compilation, lint, and smoke tests. A key innovation is a counterexample-guided formal agent that converts bounded-model-checking traces into reusable micro-tests, significantly reducing bug detection time and repair iterations. The portable escalation controller can wrap existing Verilog LLM pipelines without modifying their internals. Evaluated on VerilogEval Human, VerilogEval V2, and RTLLM benchmarks, HDLFORGE demonstrates improved accuracy-latency trade-offs compared to single-stage systems through comprehensive analysis of wall-clock time distributions, escalation thresholds, and agent ablations. On VerilogEval Human and VerilogEval V2, HDLFORGE-Qwen achieves 91.2% and 91.8% Pass@1 with roughly 50% lower median latency, dramatically improving accuracy over other medium-sized models, and 97.2% Pass@5 on RTLLM.