I built an open-source agent harness that lets a 2B model finish real tasks (288-cell benchmark, all data public)

说明与用途

Mingbird 是一个开源代理框架，使 2‑B 模型能够完成真实任务，在 18 个任务上达到 82% 的成功率，接近 35‑B 模型的表现。

问题

Small 2‑B models often fail to complete real tasks reliably, looping on tool calls or declaring victory prematurely.

结果

Mingbird enabled a 2‑B model to complete real tasks with high reliability, closing the performance gap with larger models and outperforming other harnesses on a 288‑cell benchmark.

工作流

• Run 18 real tasks with a 2‑B model using Mingbird harness
• Measure success via deterministic scoring of produced artifacts
• Compare results against other harnesses and larger models

证据

Every person who has tried to make a 2B model do real work on their own machine knows the failure mode. The model loops on the same tool call, or declares victory with half the deliverables missing, or just goes silent in the middle of the task.

来源

Mingbird handles that layer, and a 2B model finishes real work.

来源

Four agent harnesses (Mingbird, goose, opencode, agent-mini, all stock, unpatched) × four open models (gemma4:e2b 2B, qwen3.5:4b, gemma4:12b, ornith-1.5:35b) × 18 real tasks = 288 cells.

来源

Overall: 0.886 / 0.631 / 0.479 / 0.405.

来源

At 2B, three of the four harnesses fall off a cliff and Mingbird does not: 0.821 is within 0.12 of its own 35B score.

来源

One task (WF-08), same 4B model, two builds of my own harness: Before one fix: 156 tool calls, zero files produced. After: 15 tool calls, three files.

来源