yeyuxin777/baoyan-agent-benchmark

选用指南

AI 根据 README 整理 · 2026/10/15 19:31:23

使用方式

通过命令行运行，使用Python标准库，无需pip安装。需要配置官方客户端和订阅或API密钥。

查看 README 依据

Run from the repository root. Automated native execution requires macOS Seatbelt and Python 3.10+ (3.11+ for Kimi subscription configuration). The framework uses only the standard library; no `pip install` is needed. Install the official client and prepare your own subscription login or API key. Choose a configuration from the current [agent registry](config/agents-cli.example.json). These IDs are used with `--agent`; they are not client model arguments: | Agent configuration | Agent ID | Client | Template authentication | | --- | --- | --- | --- |

| Kimi K3 | `kimi-k3` | Kimi Code / `kimi` | Subscription OAuth | Every configuration uses the same workflow. Replace `AGENT` with an ID above and retain its chosen authentication mode: ```sh AGENT=gpt-6-1-sol python3 -m benchmark setup --agent "$AGENT" python3 -m benchmark cli-check --agent "$AGENT" python3 -m benchmark cli-batch \ --selection config/selections/current.json \ --agent "$AGENT" --out "./runs/current-$AGENT" ```

部署要求

需要Python 3.10+（Kimi订阅配置需3.11+），macOS Seatbelt用于自动化原生执行。

查看 README 依据

Run from the repository root. Automated native execution requires macOS Seatbelt and Python 3.10+ (3.11+ for Kimi subscription configuration). The framework uses only the standard library; no `pip install` is needed. Install the official client and prepare your own subscription login or API key. Choose a configuration from the current [agent registry](config/agents-cli.example.json). These IDs are used with `--agent`; they are not client model arguments: | Agent configuration | Agent ID | Client | Template authentication | | --- | --- | --- | --- |

已知限制

结果基于不同运行，输入版本和预算不完全统一，且并非所有公共记录都有完整工具轨迹。

查看 README 依据

| 8 | [Antigravity + Gemini 3.1 Pro](experiments/records70/models/antigravity-gemini-3-1-pro-high/README.md) | 70 | 70/70 (100.0%) | 55.55 | All eight agent configurations completed 70 tasks; 0 current tasks timed out. These results combine different real runs with input versions and budgets that are not fully uniform. Complete tool traces are not available for every public record; the leaderboard does not imply uniform execution conditions.