MetaGPT
https://github.com/FoundationAgents/MetaGPT
Cross-listing of MetaGPT as a harness. Multi-agent OSS framework that simulates a software company — PM, architect, engineer, QA. Generates docs + code + tests from a one-line prompt. Distinct from "Agent frameworks (no first-party memory layer)" row in framing.
Active · Verified 2026-10-31 · fresh as of 2026-10-31 · 47 cited facts, 1 with a third-party source · 2 recorded changes · Suggest a correction
At a glance
- Tier
- T2
- Section
- Agent IDEs & coding harnesses
- Latest release
- v0.8.1 (2024-04-22) marked Latest; v0.8.2 tag (2025-03-09) repo fix
- Pricing
- Free (OSS) repo fix
- Funding
- not applicable — OSS / no commercial entity fix
- Freshness
- 2026-10-31 fix
- Layer
- agent-infra repo fix
- Licensing
- open-source repo fix
- Developer activity
- OpenRank 1.67 (12-mo avg 1.8) · 29 participants — 2026-09 third-party fix
When to use
Optimised for: multi-agent role-play software-company simulation
Anti-fit: research-grade; not production-stable; output quality varies
Pros & cons
Pros
Most-starred multi-agent harness; defined the "agent-as-role" pattern many newer agents copy.
Cons
Research-grade output; weaker than focused coding agents (Devin/Cursor) on real-world tasks; documentation is mixed.
Claims & capabilities
One of the highest-starred multi-agent frameworks; introduced "agent-as-role" pattern.
Technical surface
- API surface
- CLI + Python SDK
- Backend storage
- local workspace + role memory
- Deployment
- Self-hosted only — Python library/CLI (pip install metagpt)
- Embedding model
- not applicable — not a memory product
- Multi-tenancy
- single-user
- MCP
- searched not found
- A2A
- searched not found
- OpenTelemetry
- searched not found
Alternatives to MetaGPT → live options ranked by shared design, licence, hosting and MCP support
Similar systems
Other agent ides & coding harnesses in the catalog, ranked by inbound references.
- Claude Code (Anthropic) T1
Anthropic's agentic coding tool: Claude as a coding agent in the terminal, IDEs (VS Code, JetBrains), desktop app, web/mobile, GitHub and Slack. Context via CLAUDE.md, skills, hooks, subagents and MCP servers.
- Cursor (IDE) T1
AI-first IDE that forked VS Code; Composer (multi-file edit) and Agent Mode (long-running coding agent) are its flagship surfaces. Memory layer = .cursorrules / Rules; recall via repo-aware Context Engine. Anysphere (parent) raised $900M Series C at $9.9B in May 2025; acquired by SpaceX (announced 2026-08-14, terms undisclosed), product continues under the Cursor brand.
- OpenAI Codex CLI T1
OpenAI's actively developed OSS terminal coding agent (Apr 2025), now written in Rust (codex-rs replaced the original TS CLI); also IDE extension + desktop app. Separate from the Codex Web cloud agent and the retired 2021 Codex model API. The codex mcp-server command (Codex as an MCP server) was deprecated on 2026-08-24 and removed in CLI 0.154.0 (2026-09-09); OpenAI points integrations to the experimental Codex app server, and codex mcp still connects Codex to external MCP servers.
- Opencode (SST) T2
SST team's open-source terminal coding agent — TypeScript, multi-model, MCP-native. Released 2025; explicitly positioned as a permissively-licensed Claude Code alternative. Memory via AGENTS.md (the emerging cross-harness convention).
- Bedrock AgentCore (AWS) T2
AWS Bedrock AgentCore — managed runtime to deploy + scale + observe agents. Multi-component: Runtime, Identity, Memory, Gateway, Browser, Code Interpreter, Observability. Distinct from Bedrock Agents (older fixed-config product); AgentCore is the 2025 framework-agnostic runtime for any agent (LangGraph, CrewAI, Strands, OpenAI Agents SDK).
- GitHub Copilot (Agent Mode) T1
The original AI coding assistant (Copilot in editor; 2021). 2025 evolved into a full agentic surface: chat, completions, Agent Mode (multi-step), and Copilot Workspace. Backed by Microsoft. Distinct row from Workspace — this row covers the everyday in-editor + agent mode product, Workspace covers the hosted spec-driven workflow.
Sources (46 cited facts, 1 third-party)
Each fact links to the page it came from; the badge says who published it (how to read sources).
- What it is Multi-agent software-company simulation (OSS) repo
- What it does & what's distinct Cross-listing of MetaGPT as a harness. Multi-agent OSS framework that simulates a softwa… repo
- Claims / benefits One of the highest-starred multi-agent frameworks; introduced "agent-as-role" pattern. repo
- Modalities text repo
- Created 2023-06 repo
- Latest release v0.8.1 (2024-04-22) marked Latest; v0.8.2 tag (2025-03-09) repo
- License MIT repo
- GitHub FoundationAgents/MetaGPT — ~71k stars, last commit n/a repo
- Mindshare 20+ HN threads; broad academic reference in multi-agent surveys repo
- Customers / scale OSS — broad community + research labs repo
- Pricing Free (OSS) repo
- Compliance OSS — no formal compliance repo
- Data handling Local CLI by default — LLM calls go through user-configured provider (OpenAI/Anthropic/A… repo
- Deployment Self-hosted only — Python library/CLI (pip install metagpt) repo
- HQ China (DeepWisdom team) repo
- Founders / pedigree Chenglin Wu (Alexander Wu), Sirui Hong et al. — DeepWisdom repo
- Memory volume / scale no scale claim repo
- Code/weights release Open-source — github.com/FoundationAgents/MetaGPT repo
- API surface CLI + Python SDK repo
- Backend storage local workspace + role memory repo
- Multi-tenancy single-user repo
- Namespace primitives per-project + per-role repo
- Contradiction handling LLM resolves repo
- Runtimes Python registry
- Validated verticals research / agent experimentation; not production SWE repo
- Time to running <30min repo
- Anti-fit research-grade; not production-stable; output quality varies repo
- Optimised for multi-agent role-play software-company simulation repo
- Adjacent infrastructure OpenAI + Anthropic + Azure OpenAI repo
- Pros Most-starred multi-agent harness; defined the "agent-as-role" pattern many newer agents … repo
- Cons Research-grade output; weaker than focused coding agents (Devin/Cursor) on real-world ta… repo
- Project & sources github.com/FoundationAgents/MetaGPT; metagpt paper repo
- commit-trajectory [{"ym":"2024-08","cum":5539},{"ym":"2024-09","cum":5887},{"ym":"2024-10","cum":6164},{"y… repo
- download-trajectory [{"ym":"2025-11","monthly":9770,"cum":9770},{"ym":"2025-12","monthly":33649,"cum":43419}… registry
- cost-input-usd-per-mtok 0 repo
- cost-output-usd-per-mtok 0 repo
- cost-tier free repo
- cost-pricing-model free-tier repo
- cost-last-verified 2026-06-29 repo
- capability-benchmark-sources humaneval: 85.9% (MetaGPT with executable feedback, pass@1, paper arXiv 2308.00352 v7 20… paper
- architecture-deployment-locus local repo
- licensing-bucket open-source repo
- positioning-platform-vs-product platform repo
- taxonomy-three-layer-rollup agent-infra repo
- developer-activity-opendigger {"activity":49.95,"as_of":"2026-09","openrank_12mo":1.8,"openrank_current":1.67,"partici… third-party
- sdk-languages python repo
What changed
Changes to this record's lifecycle, licence, release, pricing and hosting facts, from the catalog's git history. Subscribe (Atom)
Related systems
References (3)
- Anthropic Claude (foundation models) depends on at runtime — adjacent-infrastructure cell: OpenAI + Anthropic + Azure OpenAI
- Azure Machine Learning depends on at runtime — adjacent-infrastructure cell: OpenAI + Anthropic + Azure OpenAI
- OpenAI GPT family (GPT-5 / GPT-4o / o3 / o4) depends on at runtime — adjacent-infrastructure cell: OpenAI + Anthropic + Azure OpenAI