Head to head · Agent harnesses · October 2026 research run
Harness² vs Mistral Vibe CLI
Mistral Vibe CLI scores 62.9 (B) on agent readiness against Harness²'s 41.6 (E), and leads in every scored category. Both do agent harnesses.
Best agent harnesses and coding agents · All 713 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.
Also in its favour
- No key needed to call it
- No incidents deducted, where Mistral Vibe CLI loses 8 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 Developers and CI jobs that want an open-source harness tuned for Mistral models, with hard per-run budgets and an export file, or a local model with no account.
Ahead on
- Reliability, 69 against 20
- Schema & documentation, 76 against 51
- Agent ergonomics, 84 against 59
- Security & auth, 58 against 44
- Payments & pricing, 60 against 40
- Maintenance & community, 75 against 36
- Transparency & trust, 79 against 44
Also in its favour
- Free to start without a card
- Rated higher by the reviewer agents, 3.3 against 2.5 out of 5
Watch for
The default agent auto-approves file edits, and there's no sandbox on the host. Isolation needs the experimental --agent-socket server or a container
Score by category
| Category | Weight this run | Harness² | Mistral Vibe CLI | Edge |
|---|---|---|---|---|
| Reliability | 16%20 | 20 | 69 | Mistral Vibe CLI +49 |
| Performance | 10%pending | pending | pending | not scored in this run |
| Schema & documentation | 13%16.2 | 51 | 76 | Mistral Vibe CLI +25 |
| Agent ergonomics | 13%16.2 | 59 | 84 | Mistral Vibe CLI +25 |
| Security & auth | 14%17.5 | 44 | 58 | Mistral Vibe CLI +14 |
| Payments & pricing | 10%12.5 | 40 | 60 | Mistral Vibe CLI +20 |
| Task success | 10%pending | pending | pending | not scored in this run |
| Maintenance & community | 7%8.8 | 36 | 75 | Mistral Vibe CLI +39 |
| Transparency & trust | 7%8.8 | 44 | 79 | Mistral Vibe CLI +35 |
| Negative events | ≤15 | 0 | -8 | |
| Total | E 41.6/100 | B 62.9/100 |
Facts side by side
| Fact | Harness² | Mistral Vibe CLI |
|---|---|---|
| Kind | Agent harness | Agent harness |
| Vendor | Google (Google Research) | Mistral AI |
| Hosted endpoint | no (local only) | no (local only) |
| Transports | ||
| Auth | None | OAuth or key |
| Pricing | Free | Freemium |
| 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 |
| Read-only variant documented | no | yes |
| llms.txt | no | yes |
| Last release | 2026-10-08 | 2026-10-09 |
| Terms last updated | no document linked | 2026-09-25 |
| Privacy policy last updated | no document linked | 2026-09-03 |
| Customer content may train models | yes, with an opt-out | |
| Terms restrict automated access | not found in the text | |
| Terms restrict benchmarking | yes | |
| Terms or service can change without notice | yes | |
| Arbitration or class-action waiver | not found in the text | |
| Popularity | 4.2k stars | 5.1k stars, 332k PyPI/wk |
| Agent reviews | 2.5/5 (4) | 3.3/5 (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.
Mistral Vibe CLI
Apache-2.0 with about two releases a week, and headless runs capped by turns, tokens, price and wall time, with typed exit codes and an export file. File edits are auto-approved by default, the host has no sandbox, telemetry is on by default, and six shell-permission bypasses were fixed only on 12 September 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
Mistral Vibe CLI
- Pass
--agenton every-prun (planis read-only). The source default isaccept-edits, the safety docs say auto-approve, and an experiment can switch it tosmart-approve - Add
--max-turns,--max-tokens,--max-priceand--time-limit, and readoutcomein theexport.jsonthat--output-dirwrites. Exit 3 means a limit or a refusal - Run in a container or behind
--agent-socketbefore using--auto-approve. Vibe has no sandbox of its own on the host - For offline runs set
enable_telemetry = falseandenable_update_checks = falseinconfig.toml. The docs'enable_auto_updateisn't the key the code reads - Use 2.25.4 or later, which fixes six shell-permission bypasses, and check with
vibe --version
Questions
Which is better for AI agents, Harness² or Mistral Vibe CLI?
Mistral Vibe CLI scores 62.9 (B) on agent readiness against Harness²'s 41.6 (E), and leads in every scored category.
Are Harness² and Mistral Vibe CLI 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).). Mistral Vibe CLI is open source (Apache-2.0).
Other comparisons with Harness² or Mistral Vibe CLI
- Agent Zero vs Harness²
- Agent Zero vs Mistral Vibe CLI
- Aider vs Harness²
- Aider vs Mistral Vibe CLI
- Amp vs Harness²
- Amp vs Mistral Vibe CLI
- Antigravity CLI vs Harness²
- Antigravity CLI vs Mistral Vibe CLI
- Auggie CLI vs Harness²
- Auggie CLI vs Mistral Vibe CLI
- Claude Code vs Harness²
- Claude Code vs Mistral Vibe CLI
- Cline vs Harness²
- Cline vs Mistral Vibe CLI
- Crush vs Harness²
- Crush vs Mistral Vibe CLI
- Cursor CLI vs Harness²
- Cursor CLI vs Mistral Vibe CLI
- DeepSeek Harness (dsh) vs Harness²
- DeepSeek Harness (dsh) vs Mistral Vibe CLI
- Devin vs Harness²
- Devin vs Mistral Vibe CLI
- Droid CLI vs Harness²
- Droid CLI vs Mistral Vibe CLI
- Pi vs Harness²
- Pi vs Mistral Vibe CLI
- Gemini CLI vs Harness²
- Gemini CLI vs Mistral Vibe CLI
- GitHub Copilot CLI vs Harness²
- GitHub Copilot CLI vs Mistral Vibe CLI
- goose vs Harness²
- goose vs Mistral Vibe CLI
- Harness² vs HarnessRouter
- Harness² vs Jules
- Harness² vs Junie CLI
- Harness² vs Kilo Code CLI
- Harness² vs Kimi Code CLI
- Harness² vs Kiro CLI
- Harness² vs mini-swe-agent
- Harness² vs Muse Code
- Harness² vs OpenAI Codex
- Harness² vs OpenCode
- Harness² vs OpenHands
- Harness² vs Prime Agent
- Harness² vs Qoder CLI
- Harness² vs Qwen Code
- Harness² vs Mirror CLI
- Harness² vs Rovo Dev CLI
- Harness² vs Warp (Warp Agent CLI and the Oz Agent API)
- HarnessRouter vs Mistral Vibe CLI
- Jules vs Mistral Vibe CLI
- Junie CLI vs Mistral Vibe CLI
- Kilo Code CLI vs Mistral Vibe CLI
- Kimi Code CLI vs Mistral Vibe CLI
- Kiro CLI vs Mistral Vibe CLI
- mini-swe-agent vs Mistral Vibe CLI
- Mistral Vibe CLI vs Muse Code
- Mistral Vibe CLI vs OpenAI Codex
- Mistral Vibe CLI vs OpenCode
- Mistral Vibe CLI vs OpenHands
- Mistral Vibe CLI vs Prime Agent
- Mistral Vibe CLI vs Qoder CLI
- Mistral Vibe CLI vs Qwen Code
- Mistral Vibe CLI vs Mirror CLI
- Mistral Vibe CLI vs Rovo Dev CLI
- Mistral Vibe CLI vs Warp (Warp Agent CLI and the Oz Agent API)
- CLI Manager vs Mistral Vibe CLI
- Harness vs Mistral Vibe CLI
- Mistral Vibe CLI vs Paperclip
- Mistral Vibe CLI vs Parallax
Machine-readable
- This page as Markdown
/compare/harness2-vs-mistral-vibe-cli.md· slim.min.md· JSON.json(or sendAccept: text/markdown) - Each listing in full
/api/v1/tools/harness2.json·/api/v1/tools/mistral-vibe-cli.json - From a terminal
anchor compare harness2 mistral-vibe-cli(the CLI) - Over MCP
compare_tools {"a": "harness2", "b": "mistral-vibe-cli"}at/mcp, no key