---
language:
- en
- zh
license: mit
base_model: zai-org/GLM-5.2
library_name: transformers
pipeline_tag: text-generation
tags:
- macaron
- macaron-v1
- glm-5.2
- coding-agent
- merged-lora
- tool-use
- terminal-agent
- swe-agent
---

# Macaron-V1-Coding-Venti

<div align="center">
  <img src="assets/mindlab_logo.svg" width="32%" alt="MindLab logo"/>
</div>

<p align="center">
  🧭 <b>Parent model:</b> <a href="https://huggingface.co/mindlab-research/Macaron-V1-Venti" target="_blank">Macaron-V1-Venti</a>
  <br>
  📖 <b>Blog:</b> <a href="https://macaron.im/mindlab/research/introducing-macaron-v1" target="_blank">Introducing Macaron-V1</a>
  <br>
  🚀 <b>Hosted API:</b> <a href="https://mintcn.macaron.xin/" target="_blank">Macaron API Platform</a>
  <br>
  🧩 <b>Artifacts:</b> <a href="https://github.com/MindLab-Research/macaron-artifacts" target="_blank">Macaron Artifacts</a>
  <br>
  🛠️ <b>Self-hosted serving:</b> <a href="https://github.com/MindLab-Research/Mixture-of-LoRA-Harness" target="_blank">Mixture of LoRA (MoL) serving harness</a>
  <br>
  📄 <b>Technical report:</b> coming soon
</p>

Macaron-V1-Coding-Venti is the coding-specialist checkpoint in the Macaron-V1 family. It is produced by merging the Macaron-V1-Venti L2 Coding LoRA specialist into the GLM-5.2 BF16 base model, yielding a single merged checkpoint for coding, terminal, repository, and software-engineering workflows.

This repository is intended for deployments that want the coding specialist as a standard merged model without runtime LoRA routing. For the full routed Chat, Agent, Coding, and GenUI Mixture of LoRA (MoL) system, use [Macaron-V1-Venti](https://huggingface.co/mindlab-research/Macaron-V1-Venti).

## Highlights

- A merged GLM-5.2 coding-specialist checkpoint derived from the Macaron-V1-Venti L2 Coding LoRA.
- No runtime adapter loading is required: the LoRA update is merged directly into the BF16 base weights.
- Built for code understanding, repository-level software engineering, terminal use, and coding-agent workflows.
- Shares the Macaron-V1 post-training stack and evaluation harness used for the flagship Macaron-V1-Venti release.

## Model Overview

| Field | Value |
|---|---|
| Model name | Macaron-V1-Coding-Venti |
| Organization | MindLab Research |
| Release family | Macaron-V1 |
| Base model | GLM-5.2 |
| Source specialist | Macaron-V1-Venti L2 Coding LoRA |
| Checkpoint type | BF16 merged checkpoint |
| Runtime architecture | GLM-5.2 weights with coding LoRA update merged into the base |
| Primary domains | Coding, terminal workflows, software-engineering agents |
| Context length | 1M |
| License | MIT |

## Evaluation

The coding-specialist release reports the coding and terminal slices from the Macaron-V1 evaluation suite.

| Benchmark | Macaron V1 | GLM 5.2 | GPT 5.5 | Claude Opus 4.8 | Gemini 3.1 Pro | Qwen 3.7 Max | Minimax M3 |
|---|---:|---:|---:|---:|---:|---:|---:|
| SWE Verified | 85.6 | 80.4 | 82.9 | 88.6 | 80.6 | 80.4 | 80.5 |
| TerminalBench 2.1 | 87.6 | 82.7 | 83.4 | 78.9 | 70.7 | 73.5 | 66.0 |
| DeepSWE | 58.4 | 54.9 | 70.0 | 58.0 | 10.0 | 18.0 | 20.0 |
| SWE Atlas QnA | 49.5 | 48.9 | 45.4 | 57.3 | 13.5 | 22.6 | 37.9 |

Higher is better for all scores shown in the table.

### Evaluation Protocols

The coding benchmark table is mirrored in [`evaluation/coding_benchmark_summary.yaml`](evaluation/coding_benchmark_summary.yaml) for reproducibility and downstream parsing.

The full benchmark methodology will be released with the technical report.

## Usage

This repository contains a merged checkpoint. Load it as a standard GLM-5.2-compatible causal language model; no PEFT adapter attachment or MoL router is required.

```python
import torch
from transformers import AutoModelForCausalLM, AutoTokenizer

repo_id = "mindlab-research/Macaron-V1-Coding-Venti"

tokenizer = AutoTokenizer.from_pretrained(
    repo_id,
    trust_remote_code=True,
)

model = AutoModelForCausalLM.from_pretrained(
    repo_id,
    torch_dtype=torch.bfloat16,
    device_map="auto",
    trust_remote_code=True,
)
model.eval()
```

For production serving, use the same GLM-5.2-compatible serving stack you would use for a merged BF16 checkpoint. The full Macaron routed system, including Chat, Agent, Coding, and GenUI specialists, remains available through [Macaron-V1-Venti](https://huggingface.co/mindlab-research/Macaron-V1-Venti) and the [Mixture of LoRA (MoL) serving harness](https://github.com/MindLab-Research/Mixture-of-LoRA-Harness).

## Macaron Artifacts

[Macaron Artifacts](https://github.com/MindLab-Research/macaron-artifacts) is the companion local WebUI and plugin bundle for viewing Macaron sessions and generated artifacts. It supports Claude Code, Codex, and Kimi Code, and can run against Macaron or another Anthropic-compatible endpoint.

Install it as a plugin in the agent runtime you use:

```text
# Claude Code
/plugin marketplace add https://github.com/MindLab-Research/macaron-artifacts
/plugin install macaron@macaron
```

```bash
# Codex
codex plugin marketplace add https://github.com/MindLab-Research/macaron-artifacts
codex plugin add macaron@macaron
```

```text
# Kimi Code
/plugins install https://github.com/MindLab-Research/macaron-artifacts
/reload
```

## License

This repository is released under the MIT License. Users should also respect any requirements inherited from the GLM-5.2 base model and from dependencies used by their serving stack.

## Citation

```bibtex
@misc{mindlab2026macaronv1,
  author = {{Mind Lab}},
  title = {Introducing Macaron-V1},
  year = {2026},
  howpublished = {Mind Lab: A Lab for Experiential Intelligence},
  note = {https://macaron.im/mindlab/research/introducing-macaron-v1}
}
```