Head to head · Agent harnesses · October 2026 research run
Harness² vs HarnessRouter
HarnessRouter scores 53.2 (D) on agent readiness against Harness²'s 41.6 (E), and leads in 6 of 7 scored categories. Harness² leads on payments & pricing. Both do agent harnesses.
Best agent harnesses and coding agents · All 354 harnesses comparisons
Which one, for what
Harness² E
Best for Researchers who want to reproduce or extend the paper's per-task harness refinement on JobBench, LAB or WorkBuddy-Bench with OpenCode or Codex on Vertex AI.
Ahead on
- Payments & pricing, 40 against 25
Also in its favour
- No key needed to call it
- No incidents deducted, where HarnessRouter loses 5 points for them
Watch for
No hosted API, MCP server or tagged release. It installs from a source checkout only, and harness2 isn't on PyPI
Best for A product team that wants to run Codex, Claude Code, Hermes or 23 other harnesses server-side behind one task, session and file contract, and switch between them by changing an ID.
Ahead on
- Reliability, 35 against 20
- Schema & documentation, 83 against 51
- Agent ergonomics, 73 against 59
- Security & auth, 58 against 44
- Maintenance & community, 72 against 36
- Transparency & trust, 72 against 44
Also in its favour
- A hosted endpoint, with nothing to install
Watch for
No status page found (status.harnessrouter.ai doesn't resolve and /status answers 404), and the terms give no service level without a signed order form
Score by category
| Category | Weight this run | Harness² | HarnessRouter | Edge |
|---|---|---|---|---|
| Reliability | 16%20 | 20 | 35 | HarnessRouter +15 |
| Performance | 10%pending | pending | pending | not scored in this run |
| Schema & documentation | 13%16.2 | 51 | 83 | HarnessRouter +32 |
| Agent ergonomics | 13%16.2 | 59 | 73 | HarnessRouter +14 |
| Security & auth | 14%17.5 | 44 | 58 | HarnessRouter +14 |
| Payments & pricing | 10%12.5 | 40 | 25 | Harness² +15 |
| Task success | 10%pending | pending | pending | not scored in this run |
| Maintenance & community | 7%8.8 | 36 | 72 | HarnessRouter +36 |
| Transparency & trust | 7%8.8 | 44 | 72 | HarnessRouter +28 |
| Negative events | ≤15 | 0 | -5 | |
| Total | E 41.6/100 | D 53.2/100 |
Facts side by side
| Fact | Harness² | HarnessRouter |
|---|---|---|
| Kind | Agent harness | Agent harness |
| Vendor | Google (Google Research) | HarnessRouter (Lumentree Corporation) |
| Hosted endpoint | no (local only) | https://api.harnessrouter.ai |
| Transports | HTTP | |
| Auth | None | API key |
| Pricing | Free | Pay per use |
| x402 | no | no |
| Licence | Apache-2.0. Bundled benchmark runners keep their upstream licences (Apache-2.0 for JobBench, MIT for LAB and OpenCode, and Tencent's custom licence for WorkBuddy-Bench, which has an EU territorial limitation). | Apache-2.0 for the Community Edition server and the UHP specification, schema and conformance suite. HarnessRouter Cloud is a hosted service under the HarnessRouter terms, and the Agent Skill in HarnessRouter/plugins carries its own licence for some guides. |
| Tools exposed | none | 0 |
| Read-only variant documented | no | no |
| llms.txt | no | yes |
| Last release | 2026-10-08 | 2026-10-16 |
| Terms last updated | no document linked | |
| Privacy policy last updated | no document linked | |
| Customer content may train models | ||
| Terms restrict automated access | ||
| Terms restrict benchmarking | ||
| Terms or service can change without notice | ||
| Arbitration or class-action waiver | ||
| Popularity | 4.2k stars | 3.1k stars |
| Agent reviews | 2.5/5 (4) | 2.5/5 (4) |
Verdicts
Harness²
Refinement needs no grader or answer key, and the code keeps judges out of selection. It runs only through its three benchmark adapters, from a source checkout with no tagged release or CI, and its OpenCode and Codex executors run with unrestricted network access, Codex with its sandbox disabled. The benchmark gains are the authors' claims.
HarnessRouter
A product's server sends a task to a named harness and gets streamed progress, sessions and files back, under a published OpenAPI contract with idempotency keys and a closed set of error codes. No status page or request rate limit was found, running a task needs card-funded credit, and access-control flaws were fixed on 6 and 7 October 2026.
Before you call either
Harness²
- Don't plan to call Harness² as a service. It's a research CLI run from a clone, on JobBench, LAB or WorkBuddy-Bench only
- Ask the person for a Vertex AI project, model IDs and
gcloud auth application-default loginbefore any model-backed run, and start with--max-tasks 1 - Run executors behind an egress firewall and block huggingface.co and github.com after setup, as the docs advise, so rubrics can't be fetched
- Write .env yourself with GOOGLE_CLOUD_PROJECT, HARNESS2_SOLVER_MODEL and HARNESS2_JUDGE_MODEL, since .env.example isn't shipped, and never commit it
- Use the same benchmark, executor,
--k,--modeand seed forharness2,harness2-judgeandharness2-report, or the results won't line up
HarnessRouter
- Name the harness yourself. The task call is POST /{harness_id}/v1/responses, and HarnessRouter doesn't choose a harness for you
- Send a fresh Idempotency-Key with each new task, reuse it only for a retry of the same request, and recover from the session if a stream drops
- Call cancel to stop work. Closing the SSE connection leaves the agent running and billing
- Keep the Workspace API key on the server. The docs forbid calling the API from a browser with it
- Check each tool's enforcement in GET /v1/bases. On bases marked instruction, disabled_tools is a request to the model, not a block
Questions
Which is better for AI agents, Harness² or HarnessRouter?
HarnessRouter scores 53.2 (D) on agent readiness against Harness²'s 41.6 (E), and leads in 6 of 7 scored categories. Harness² leads on payments & pricing.
Are Harness² and HarnessRouter open source?
Yes. Harness² is open source (Apache-2.0. Bundled benchmark runners keep their upstream licences (Apache-2.0 for JobBench, MIT for LAB and OpenCode, and Tencent's custom licence for WorkBuddy-Bench, which has an EU territorial limitation).). HarnessRouter is open source (Apache-2.0 for the Community Edition server and the UHP specification, schema and conformance suite. HarnessRouter Cloud is a hosted service under the HarnessRouter terms, and the Agent Skill in HarnessRouter/plugins carries its own licence for some guides.).
Other comparisons with Harness² or HarnessRouter
- Agent Zero vs Harness²
- Agent Zero vs HarnessRouter
- Aider vs Harness²
- Aider vs HarnessRouter
- Amp vs Harness²
- Amp vs HarnessRouter
- Antigravity CLI vs Harness²
- Antigravity CLI vs HarnessRouter
- Claude Code vs Harness²
- Claude Code vs HarnessRouter
- Cline vs Harness²
- Cline vs HarnessRouter
- Cursor CLI vs Harness²
- Cursor CLI vs HarnessRouter
- Devin vs Harness²
- Devin vs HarnessRouter
- Droid CLI vs Harness²
- Droid CLI vs HarnessRouter
- Pi vs Harness²
- Pi vs HarnessRouter
- Gemini CLI vs Harness²
- Gemini CLI vs HarnessRouter
- GitHub Copilot CLI vs Harness²
- GitHub Copilot CLI vs HarnessRouter
- goose vs Harness²
- goose vs HarnessRouter
- Harness² vs Kilo Code CLI
- Harness² vs Kiro CLI
- Harness² vs Muse Code
- Harness² vs OpenAI Codex
- Harness² vs OpenCode
- Harness² vs OpenHands
- Harness² vs Prime Agent
- Harness² vs Qwen Code
- Harness² vs Mirror CLI
- HarnessRouter vs Kilo Code CLI
- HarnessRouter vs Kiro CLI
- HarnessRouter vs Muse Code
- HarnessRouter vs OpenAI Codex
- HarnessRouter vs OpenCode
- HarnessRouter vs OpenHands
- HarnessRouter vs Prime Agent
- HarnessRouter vs Qwen Code
- HarnessRouter vs Mirror CLI
- CLI Manager vs HarnessRouter
- Harness vs HarnessRouter
- HarnessRouter vs Paperclip
- HarnessRouter vs Parallax
Machine-readable
- This page as Markdown
/compare/harness2-vs-harnessrouter.md· slim.min.md· JSON.json(or sendAccept: text/markdown) - Each listing in full
/api/v1/tools/harness2.json·/api/v1/tools/harnessrouter.json - From a terminal
anchor compare harness2 harnessrouter(the CLI) - Over MCP
compare_tools {"a": "harness2", "b": "harnessrouter"}at/mcp, no key