{"data":{"slug":"uncsoft-anubis-oss","name":"anubis-oss","tagline":"Local LLM Testing & Benchmarking for Apple Silicon","github_url":"https://github.com/uncSoft/anubis-oss","owner":"uncSoft","repo":"anubis-oss","owner_avatar_url":"https://avatars.githubusercontent.com/u/181050313?v=4","primary_language":"Swift","stars":207,"forks":15,"topics":["apple-silicon","benchmarking","gpu","inference","llm","local-llm","macos","mlx","native","ollama","performance","swiftui"],"archived":false,"github_pushed_at":"2026-09-05T00:10:19+00:00","maintenance_label":"Active","stars_delta_30d":9,"url":"https://www.graphcanon.com/tools/uncsoft-anubis-oss","markdown_url":"https://www.graphcanon.com/tools/uncsoft-anubis-oss.md","api_url":"https://www.graphcanon.com/api/graphcanon/tools/uncsoft-anubis-oss","graph_url":"https://www.graphcanon.com/api/graphcanon/graph?tool=uncsoft-anubis-oss","description":"Local LLM Testing & Benchmarking for Apple Silicon","homepage_url":"https://devpadapp.com/leaderboard.html","license":"GPL-3.0","open_issues":1,"watchers":2,"ai_summary":"Anubis-oss offers tools and benchmarks for testing and evaluating local large language models on Apple Silicon devices using Swift.","readme_excerpt":"### Hardware Stress Testing *(New in 2.9)*\n\nPush your Apple Silicon to its limits and observe power draw, thermal throttling, and frequency scaling under controlled load - all from within the Monitor.\n\n- **CPU stress** - spawns `yes` processes per core. Choose All Cores, P-Cores only, E-Cores only, or Single Core\n- **GPU stress** - Metal compute shader renders a Mandelbrot fractal zoom in a separate window. Randomized zoom targets and color palettes on each run. Four intensity levels (Low / Medium / High / Extreme) control iterations, supersampling, and passes per frame\n- **Memory bandwidth stress** - allocates memory then continuously streams through it with `memcpy` to saturate the memory bus. Reports measured bandwidth in GB/s, directly comparable to your chip's theoretical max. Three pressure levels (Light 25% / Moderate 50% / Heavy 75% of free memory)\n- **Safety mechanisms** - 5-minute auto-timeout, thermal watchdog (auto-stop at critical), GPU auto-downgrade if FPS drops below 5, cleanup on view disappear and app quit\n<img width=\"900\" alt=\"Screenshot 2026-05-28 at 4 21 50 PM\" src=\"https://github.com/user-attachments/assets/7c917eaa-b5a1-4d3b-bdc0-8e6fa339d0c7\" />\n\n---\n\n## Hardware Metrics\n\nAnubis captures Apple Silicon telemetry during inference via IOReport and system APIs:\n\n| Metric | Source | Description |\n|--------|--------|-------------|\n| GPU Utilization | IOReport | GPU active residency percentage |\n| CPU Utilization | `host_processor_info` | Usage across all cores |\n| GPU Power | IOReport Energy Model | GPU power consumption in watts |\n| CPU Power | IOReport Energy Model | CPU (E-cores + P-cores) power in watts |\n| ANE Power | IOReport Energy Model | Neural Engine power consumption |\n| DRAM Power | IOReport Energy Model | Memory subsystem power |\n| GPU Frequency | IOReport GPU Stats | Weighted average from P-state residency |\n| Process Memory | `proc_pid_rusage` | Backend process `phys_footprint` (includes Metal/GPU allocations) |\n| Thermal State | `ProcessInfo.thermalState` | System thermal pressure level |\n\n---\n\n## Requirements\n\n- **macOS 15.0** (Sequoia) or later\n- **Apple Silicon** (M1 / M2 / M3 / M4 / M5 +) - Intel is not supported\n- **8 GB** unified memory minimum (16 GB+ recommended for larger models)\n- At least one inference backend installed (Ollama recommended)\n\n---\n\n---\n\n### 1. Install Ollama (or another backend)\n\n```bash\n\n---\n\n# macOS - install Ollama\nbrew install ollama\n\n---\n\n### 2. Install Anubis\n\nThe easiest path is **Homebrew** — installs the signed, notarized .app from the latest GitHub release:\n\n```bash\nbrew install --cask uncsoft/anubis/anubis-oss\n```\n\nOr download the zip directly from the [Releases page](https://github.com/uncSoft/anubis-oss/releases) and drag to `/Applications`.\n\nAnubis auto-updates via Sparkle in either case. Once installed, Anubis will auto-detect Ollama on launch. Other backends can be added in Settings.\n\n#### Updating to a new version\n\nThe app **updates itself via Sparkle** — just open Anubis and accept the update prompt (or **Settings → Check for Updates**). That's the recommended path and works for every install method.\n\nTo update through Homebrew instead:\n\n```bash\nbrew update\nbrew upgrade --cask anubis-oss --greedy\n```\n\n> The `--greedy` flag is required because the cask is marked `auto_updates` (the app updates itself), so a plain `brew upgrade` intentionally skips it.\n\n<details>\n<summary>Homebrew says <code>Cask 'anubis-oss' is unreadable: syntax errors</code> / shows <code>&lt;&lt;&lt;&lt;&lt;&lt;&lt;</code> conflict markers</summary>\n\nThis only happens if your local tap checkout has uncommitted edits or a stash that conflicts with an update. Reset it to the published cask:\n\n```bash\ncd \"$(brew --repository)/Library/Taps/uncsoft/homebrew-anubis\"\ngit fetch origin && git reset --hard origin/main\nbrew upgrade --cask anubis-oss --greedy\n```\n</details>\n\n#### Or build from source\n\n```bash\ngit clone https://github.com/uncSoft/anubis-oss.git\ncd anubis-oss/anubis\nopen anubis.x","github_created_at":"2026-02-07T23:09:03+00:00","created_at":"2026-07-15T11:01:01.285587+00:00","updated_at":"2026-09-20T05:06:39.766107+00:00","categories":[{"slug":"evaluation-observability","name":"Evaluation & Observability","url":"https://www.graphcanon.com/categories/evaluation-observability","markdown_url":"https://www.graphcanon.com/categories/evaluation-observability.md","api_url":"https://www.graphcanon.com/api/graphcanon/categories/evaluation-observability"},{"slug":"inference-serving","name":"Inference & Serving","url":"https://www.graphcanon.com/categories/inference-serving","markdown_url":"https://www.graphcanon.com/categories/inference-serving.md","api_url":"https://www.graphcanon.com/api/graphcanon/categories/inference-serving"}],"tags":[{"slug":"apple-silicon","name":"apple-silicon"},{"slug":"benchmarking","name":"benchmarking"},{"slug":"gpu","name":"gpu"},{"slug":"inference","name":"inference"},{"slug":"llm","name":"llm"},{"slug":"local-llm","name":"local-llm"},{"slug":"macos","name":"macos"},{"slug":"mlx","name":"mlx"}],"trust":{"provenance":{"is_fork":false,"github_id":1152467476,"owner_type":"User","methodology":"github_public_v1","parent_repo":null,"near_duplicate_slugs":[]},"computed_at":"2026-09-20T05:06:37.739Z","maintenance":{"label":"Active","score":82,"methodology":"github_public_v1","releases_90d":1,"days_since_push":15,"last_release_at":"2026-08-23T01:11:03Z","stars_delta_30d":9,"open_issues_delta_30d":-3},"security_summary":{"status":"no_lockfile","scanner":null,"low_count":0,"high_count":0,"last_scan_at":"2026-07-15T11:01:02.542Z","medium_count":0,"scan_profile":"none","critical_count":0}},"capability_facts":{"scan":{"source":"repo_scan","observed_at":"2026-09-20T05:06:38.725Z"},"languages":{"value":["swift"],"source":"github.language","observed_at":"2026-09-20T05:06:38.725Z"},"license_spdx":{"value":"GPL-3.0","source":"github.license","observed_at":"2026-09-20T05:06:38.725Z"}},"decision_facts":{"hosting":null,"pricing":{"model":"freemium","summary":"The tool is free and open-source with no monetary costs for usage or distribution."},"requirements":{"min_ram_gb":8,"requires_docker":false},"constraints":{"min_ram_gb":8,"pricing_model":"freemium","requires_docker":false},"when_to_use":["When developing and evaluating large language models intended to run natively on Apple Silicon hardware.","If you require benchmarks for GPU performance specific to Mac devices."],"when_not_to_use":["If your development does not involve Apple Silicon or macOS environments as Anubis-oss is tightly integrated with these platforms.","When preferring a language other than Swift, since Anubis-oss depends on this for its operations."],"source":"enrich:decision_facts","observed_at":"2026-07-17T10:22:58.195Z"},"constraint_facets":{"min_ram_gb":8,"pricing_model":"freemium","requires_docker":false},"decision_summary":[{"label":"Pricing","value":"freemium - The tool is free and open-source with no monetary costs for usage or distribution."},{"label":"Requirements","value":"Min 8 GB RAM"},{"label":"Adopt for","value":"Anubis-oss, specifically tailored for Apple Silicon devices using Swift, is distinguished by its focus on local large language model evaluation and testing within the macOS environment."},{"label":"License detail","value":"GPL-3.0 license ensures that any derivative works related to anubis-oss must also be open source under the same licensing terms."}]}}