覆盖 459 类工作流,通过流量估算、产品功能与替代方案辅助选型。
“parallel experiments” · 编程与软件开发 · AI 代码生成器·18 个结果
重新开始


覆盖整个软件开发周期的智能体编排平台

在一个订阅中并排对比 ChatGPT、Claude、Gemini 及 30 多款 AI 模型

让 AI 编程模型同台竞技
让 AI 智能体永不休眠的持久化虚拟机。

Termo is a platform for running AI agents on their own virtual machines. Each agent keeps a semantic memory graph, can spawn sub-agents, install skills from a large registry, browse the web, execute Python, and run on a cron schedule you set in plain language — so work keeps happening while you are offline.
Termo is a platform for running AI agents on their own virtual machines. Each agent keeps a semantic memory graph, can spawn sub-agents, install skills from a large registry, browse the web, execute Python, and run on a cron schedule you set in plain language — so work keeps happening while you are offline.

Replicas is a cloud coding-agent platform for engineering teams. It runs agents like Claude Code, Codex and Cursor in isolated Linux VMs and returns pull requests — started from Slack, Linear, GitHub, GitLab or a web dashboard.

PapertLab is an open-source AI pair programmer you install with pip and run in your terminal. It collaborates with large language models to edit code across your local Git repository, with modes for asking questions, pair programming and semi-autonomous coding.
PapertLab is an open-source AI pair programmer you install with pip and run in your terminal. It collaborates with large language models to edit code across your local Git repository, with modes for asking questions, pair programming and semi-autonomous coding.

Solid is an AI agent platform that equips autonomous agents with their own machines, accounts, and spending budgets, letting them connect to the tools a job needs, build missing integrations, and hand back a verified result. Users describe a job in plain language, set access and approval rules, and the agents handle setup, execution, and quality checks.
Solid is an AI agent platform that equips autonomous agents with their own machines, accounts, and spending budgets, letting them connect to the tools a job needs, build missing integrations, and hand back a verified result. Users describe a job in plain language, set access and approval rules, and the agents handle setup, execution, and quality checks.

可与 Claude Code、Codex 和 Cursor 协同工作的前沿编程智能体

分布式性能测试的统一编排控制平面