{
  "schema_version": "2.0",
  "slug": "cactuscompute",
  "name": "Cactus",
  "agent_url": "https://cactuscompute.com",
  "category": "Infrastructure",
  "run_id": "run-rpub-v2-cactuscompute-2026-08-21",
  "run_at": "2026-08-21T08:45:00Z",
  "editor": "Hlido Editor",
  "editorial_method": "public-surface-tier-1+editorial-narrative-v2",
  "methodology_version": "2026.05",
  "methodology_url": "/methodology/public-surface-tier-1/",
  "score": 73,
  "tier": "STEADY",
  "laddoo_score": 73,
  "confidence": "medium",
  "hlido_opinion": {
    "headline": "On-device AI for phones, wearables and microcontrollers, with a cloud fallback — a focused, credible edge-inference stack (runtime + a 14MB agentic model) with 5.9k+ stars.",
    "body": "Cactus is an edge/on-device inference stack with three clearly separated pieces: Cactus Hybrid (post-trained models that know when they're wrong and escalate to the cloud), Cactus Needle (a 14MB agentic LLM doing tool-calling, device use and structured extraction on tiny devices), and Cactus Engine (a resource-constrained runtime with quantization tuned for battery, speed and memory). The positioning is sharp and technically coherent — on-device AI for phones, wearables, robots, home assistants and microcontrollers, with cloud fallback rather than pure-local dogma — and 5.9k+ GitHub stars plus visible docs, blog, a changelog and a compare page signal a real, actively developed project. For agent builders, a 14MB model that does tool-calling and structured extraction locally is a genuinely interesting primitive. The limits of a homepage review apply: the actual quantization quality, inference speed, battery claims and the 'knows when it's wrong' hybrid-escalation behaviour are exactly the things that decide whether an edge stack is usable, and none can be verified without hands-on benchmarking. Promising and well-scoped; verify the performance claims on your target hardware.",
    "voice": "Hlido Editor",
    "as_of": "2026-08-21",
    "editor_signature_pending": true
  },
  "tier_rationale": "STEADY (73) for a focused, credible edge-inference stack with a clear three-part architecture (Hybrid / Needle / Engine), a differentiated 14MB agentic model, and real signals of active development (5.9k+ stars, docs, changelog, compare page). Not higher because the load-bearing claims — quantization quality, battery/speed, and the hybrid 'knows when it's wrong' escalation — are unverifiable from the public surface and are precisely what determine on-device usability. Not FADING because the product is concrete, technically coherent and clearly maintained.",
  "what_it_does_well": [
    "Sharp, coherent positioning: on-device AI for phones/wearables/robots/microcontrollers with cloud fallback (not pure-local dogma)",
    "Cactus Needle — a 14MB agentic LLM doing tool-calling, device use and structured extraction on tiny devices — is a genuinely interesting edge primitive",
    "Clear three-part separation (Hybrid models, Needle model, Engine runtime) that maps to real deployment decisions",
    "Signals of an active, real project: 5.9k+ GitHub stars, docs, blog, changelog and a compare page"
  ],
  "what_it_fails_at": [
    "The decisive claims — quantization quality, inference speed, battery consumption — are asserted ('SOTA') but unverifiable from the homepage",
    "The Hybrid 'models that know when they're wrong and request cloud help' behaviour is unproven on the surface and hard to guarantee",
    "Edge deployment success is highly hardware-dependent; homepage says nothing about supported chips/OS matrix",
    "No pricing/licensing detail captured on the surface for the commercial pieces"
  ],
  "best_for": [
    "Mobile/embedded developers who need local inference with an optional cloud fallback",
    "Agent builders wanting a tiny on-device model that can tool-call and extract structured output offline",
    "Products with privacy, latency or connectivity constraints that rule out cloud-only inference"
  ],
  "not_recommended_for": [
    "Teams that need certified performance/battery numbers before adoption (benchmark on target hardware first)",
    "Server-side/cloud-only workloads where on-device constraints add no value",
    "Anyone needing broad, documented hardware-compatibility guarantees up front"
  ],
  "red_flags": [
    "Edge-inference value lives entirely in real-world quantization quality, speed and battery on YOUR hardware — these 'SOTA' claims can't be verified from the site, so benchmark before committing.",
    "The Hybrid self-doubt-and-escalate mechanism is a strong claim with no surface evidence of how reliable it is."
  ],
  "compared_to": [
    {
      "slug": "ollama",
      "verdict_diff": "Ollama makes it trivial to run open models locally on desktops/servers; Cactus targets the harder edge tier — phones, wearables, microcontrollers — with a purpose-built runtime, aggressive quantization and a tiny agentic model plus cloud fallback. Choose Ollama for easy local desktop inference, Cactus for genuinely on-device/embedded deployment.",
      "preferred_for_axis": "embedded-edge-inference-vs-desktop-local-inference"
    }
  ],
  "evidence_urls": [
    {
      "claim": "On-device AI with cloud fallback for phones, wearables, robots, home assistants and microcontrollers",
      "source": "https://cactuscompute.com (homepage hero)",
      "tested_at": "2026-08-21",
      "verified": true
    },
    {
      "claim": "Cactus Needle: a 14MB agentic LLM with tool calling, device use and structured extraction",
      "source": "https://cactuscompute.com ('Cactus Needle' product block + [NEW] banner)",
      "tested_at": "2026-08-21",
      "verified": true
    },
    {
      "claim": "Cactus Hybrid: post-trained models that know when they're wrong and request cloud help",
      "source": "https://cactuscompute.com ('Cactus Hybrid' product block)",
      "tested_at": "2026-08-21",
      "verified": true
    },
    {
      "claim": "Cactus Engine: resource-constrained inference runtime with quantization for battery/speed",
      "source": "https://cactuscompute.com ('Cactus Engine' product block)",
      "tested_at": "2026-08-21",
      "verified": true
    },
    {
      "claim": "5.9k+ GitHub stars; docs, blog, changelog and compare pages present",
      "source": "https://cactuscompute.com (hero stats + nav/footer)",
      "tested_at": "2026-08-21",
      "verified": true
    }
  ],
  "agent_relevance": {
    "has_api": true,
    "has_cli": false,
    "has_mcp": false,
    "has_webhook": false,
    "has_sdk": true,
    "behavioral_testable": true,
    "agent_integration_path": "Provides an on-device inference runtime (Cactus Engine) and models (Needle/Hybrid) with SDK/docs for integrating local tool-calling and structured extraction into apps and agents. Behaviour is testable via the open-source runtime, though real performance is hardware-dependent. Relevant to agents as an edge-inference primitive rather than a hosted agent surface.",
    "agent_friendly_score": 7
  },
  "checklist": [
    {
      "id": "homepage_loads",
      "pass": true,
      "required": true,
      "tested_at": "2026-08-21T00:00:00Z"
    },
    {
      "id": "primary_value_prop",
      "pass": true,
      "required": true,
      "evidence": "On-device AI for phones, wearables and microcontrollers, with a cloud fallback —",
      "tested_at": "2026-08-21T00:00:00Z"
    },
    {
      "id": "cta_present",
      "pass": true,
      "required": false,
      "tested_at": "2026-08-21T00:00:00Z"
    },
    {
      "id": "evidence_or_demo",
      "pass": true,
      "required": false,
      "evidence": "4 screenshot(s) captured",
      "tested_at": "2026-08-21T00:00:00Z"
    }
  ],
  "summary": "On-device AI for phones, wearables and microcontrollers, with a cloud fallback — a focused, credible edge-inference stack (runtime + a 14MB agentic model) with 5.9k+ stars.",
  "_summary_deprecation_note": "Field kept as a v1-compatibility alias of hlido_opinion.headline. New consumers should read hlido_opinion.{headline,body,voice,as_of}.",
  "staleness_after": "2026-11-21",
  "review_age_days_at_publish": 0,
  "next_review_due_at": "2026-11-21",
  "attestation_url": "/data/attestations/cactuscompute.json",
  "signature_pending": true,
  "source": "r-publish-editorial-v2",
  "marking_signal": {
    "checked_at": "2026-08-21",
    "source": "r-publish-editorial-enrich",
    "marking_statement": null,
    "detection_tool": null,
    "cop_signatory": null,
    "evidence_url": null,
    "note": "Cactus is an inference runtime/model provider; whether Article-50 output-marking applies depends on the downstream app. No marking statement or detection tool is present on the surface — recorded as unevidenced."
  },
  "evidence_images": {
    "run_id": "run-9f5b6c1990ff7b1d-cactuscompute-com",
    "base": "https://images.hlido.eu/reviews/cactuscompute/run-9f5b6c1990ff7b1d-cactuscompute-com",
    "files": [
      "home.png",
      "page_.png",
      "page_hybrid.png",
      "page_needle.png"
    ],
    "urls": [
      "https://images.hlido.eu/reviews/cactuscompute/run-9f5b6c1990ff7b1d-cactuscompute-com/home.png",
      "https://images.hlido.eu/reviews/cactuscompute/run-9f5b6c1990ff7b1d-cactuscompute-com/page_.png",
      "https://images.hlido.eu/reviews/cactuscompute/run-9f5b6c1990ff7b1d-cactuscompute-com/page_hybrid.png",
      "https://images.hlido.eu/reviews/cactuscompute/run-9f5b6c1990ff7b1d-cactuscompute-com/page_needle.png"
    ],
    "note": "Screenshots captured by the Hlido engine during the reviewed run, served from R2. `run_id` is the ENGINE run id — it differs from `scorecard.run_id` and is the only one these keys resolve under. HEAD-verification of each URL is deferred to the publish-side link-review-evidence.mjs check (this editorial enrich ran in an egress-restricted container and did not assert verification it could not perform)."
  }
}