{ "schema": "tesign-radar/item/1", "generated_at": "2026-09-17T01:56:13.055Z", "lang": "en", "slug": "slotstream", "url": "https://tesign.com/en/item/slotstream/", "alternates": { "ko": { "html": "https://tesign.com/item/slotstream/", "json": "https://tesign.com/item/slotstream/index.json", "markdown": "https://tesign.com/item/slotstream/index.md" }, "en": { "html": "https://tesign.com/en/item/slotstream/", "json": "https://tesign.com/en/item/slotstream/index.json", "markdown": "https://tesign.com/en/item/slotstream/index.md" } }, "name": "Slotstream", "use_label": "Run a 105 GB model on a 48 GB Mac", "one_liner": "Run a 105 GB Qwen model on a 48 GB Mac: experts stream from SSD, Ollama-compatible API", "text": { "why": "Qwen3.8-Flash-Next is a 105 GB mixture-of-experts model (per the README); Slotstream keeps the shared weights in RAM and reads only the experts it needs from SSD into a cache. The developer's 48 GB M5 Pro measured 13.47 tok/s after warm-up (a 0.2.16 pre-release benchmark), while a community report on a 16 GB M2 Mac mini was 1.41 tok/s. After a one-time download it works offline with no Python or cloud account.", "take": "Streaming a model larger than RAM from SSD really works here, but it supports exactly one model and speed varies widely by machine.", "build": [ "Point Open WebUI or an agent such as Hermes or fx at its OpenAI/Ollama-compatible API and you have a personal AI server on one Mac that answers questions about files and pictures without an internet connection. A Swift library lets you embed it in a Mac app directly." ], "who": [ { "role": "Apple Silicon Mac owners", "situation": "run a large model locally without a cloud account" }, { "role": "Privacy-conscious users", "situation": "keep document and image questions on the device" }, { "role": "Local-LLM developers", "situation": "study an MLX + Swift expert-streaming implementation" } ], "start": [ { "command": false, "text": "curl -fsSL https://raw.githubusercontent.com/carloslfu/slotstream/main/install.sh | sh" }, { "command": false, "text": "slotstream doctor" }, { "command": false, "text": "slotstream run --prompt \"Why is the sky blue?\"" } ], "caveat": [ "MIT for the code; model weights carry a separate Qwen community licence.", "Requires an Apple Silicon Mac with 16 GB+ RAM, macOS 14+ and about 110 GB of free SSD; 8 GB Macs refuse to start, and Windows, Linux and Intel Macs are not supported.", "The model download can take hours, long prompts can take minutes before the first token, and only one generation runs at a time." ] }, "cta": { "type": "install", "label": "INSTALL", "url": "https://sevrahq.com/" }, "ready": "install", "categories": [ "ai" ], "categories_confirmed": true, "tags": [ "local-llm", "apple-silicon", "mixture-of-experts", "mlx", "ollama", "swift" ], "platform": [ "macos", "cli" ], "license": { "spdx": "MIT", "scope": "permissive", "scope_text": "Use, change and redistribute, commercially too. Keep the notice.", "url": "https://spdx.org/licenses/MIT.html", "note": null, "open_source": true }, "language": "Swift", "source_urls": { "canonical": "https://sevrahq.com/", "github": "https://github.com/carloslfu/slotstream", "huggingface": null, "app_store": null, "sightings": [ { "source": "showhn", "url": "https://github.com/carloslfu/slotstream", "posted_at": "2026-08-31T14:46:54.000Z", "hn_url": "https://news.ycombinator.com/item?id=49510441" }, { "source": "showhn", "url": "https://github.com/carloslfu/slotstream", "posted_at": "2026-09-01T16:42:46.000Z", "hn_url": "https://news.ycombinator.com/item?id=49524447" } ] }, "numbers": { "as_of": "2026-09-16T21:00:00.000Z", "stars": 360, "stars_unit": "GitHub stars", "stars_checked_at": "2026-09-15T19:35:25.910Z", "stars_observed_since_check": 0, "star_delta_24h": null, "star_delta_7d": null, "star_delta_30d": null, "star_delta_unit": "GitHub stars gained in the window, GH Archive events summed to as_of", "spark_14d": [ null, null, null, null, null, null, null, null, null, null, null, null, null, null ], "hf_likes": null, "hf_likes_observed_at": null, "hn_points": 240, "hn_points_observed_at": "2026-09-13T13:59:17.691Z", "store": null, "method": "Star total = the value last checked on GitHub (stars_checked_at) + increases observed via GH Archive since. The 24h · 7d · 30d gains are GH Archive hourly events summed to the reference time (as_of). No score of ours." }, "ranks": null, "activity": { "pushed_at": "2026-09-15T15:01:19.000Z", "owner_type": "User", "open_issues": 7, "contributors": 3, "releases_count": 26, "latest_release": { "tag": "v0.2.18", "published_at": "2026-09-14T23:52:30.000Z", "prerelease": false }, "checked_at": "2026-09-15T19:35:25.910Z", "core_checked_at": "2026-09-15T19:35:25.910Z" }, "signals": [ { "kind": "back", "label": "BACK CATALOG", "value": null } ], "images": { "og": "https://tesign.com/img/og/slotstream.en.png", "cover": "https://tesign.com/img/slotstream-00175a10de.png", "cover_width": 1200, "cover_height": 600, "gallery": [ "https://tesign.com/img/slotstream-00175a10de.png" ], "alt": "Terminal running slotstream doctor and slotstream run: the memory-plan check and a model reply", "tier": 4 }, "dates": { "first_seen_at": "2026-09-13T13:59:17.691Z", "source_created_at": "2026-09-01T16:42:46.000Z", "discovery_hours": 285.2754697222222, "launch_signal_at": "2026-08-31T14:46:54.000Z", "kept_at": "2026-09-14T19:27:43.653Z", "first_published_at": "2026-09-14T19:27:43.752Z", "updated_at": "2026-09-15T04:36:56.999Z", "ingestion_mode": "backfill" }, "notes": [ "This file was produced by the same build, from the same data, as the tesign.com item page. The text is editorial; the numbers are stored observations.", "null in the JSON means not observed — never zero. The Markdown writes [unconfirmed] for it.", "Star total = the value last checked on GitHub (stars_checked_at) + increases observed via GH Archive since. The 24h · 7d · 30d gains are GH Archive hourly events summed to the reference time (as_of). No score of ours.", "Reconstructed from archive data, not a live discovery.", "A summary, not legal advice." ] }