{"apiVersion":"v1","taxonomyVersion":1,"generatedAt":"2026-10-29T15:46:20.179Z","data":{"recordId":"wsa_i2agxxbrk2w8","score":0.971,"confidence":"high","capped":false,"capReason":null,"sources":[{"n":1,"url":"https://github.com/ggml-org/llama.cpp/releases/tag/v0.6.0","publisher":"GitHub watchlist","authority":"owner-controlled","verification":"verified","lastRetrievedAt":"2026-10-28T00:51:38.871Z","detectedPublishedAt":"2026-10-05T16:58:23.000Z","rights":"original"},{"n":2,"url":"https://github.com/ggml-org/llama.cpp/releases/tag/b11507","publisher":"GitHub watchlist","authority":"owner-controlled","verification":"verified","lastRetrievedAt":"2026-10-08T15:33:15.947Z","detectedPublishedAt":"2026-10-08T15:31:21.000Z","rights":"original"},{"n":3,"url":"https://github.com/ggml-org/llama.cpp/releases/tag/b11559","publisher":"GitHub watchlist","authority":"owner-controlled","verification":"verified","lastRetrievedAt":"2026-10-11T22:00:32.383Z","detectedPublishedAt":"2026-10-11T19:17:37.000Z","rights":"original"},{"n":4,"url":"https://github.com/ggml-org/llama.cpp/releases/tag/b11695","publisher":"GitHub watchlist","authority":"owner-controlled","verification":"verified","lastRetrievedAt":"2026-10-16T16:41:37.619Z","detectedPublishedAt":"2026-10-16T11:16:03.000Z","rights":"original"},{"n":5,"url":"https://github.com/ggml-org/llama.cpp/releases/tag/b11714","publisher":"GitHub watchlist","authority":"owner-controlled","verification":"verified","lastRetrievedAt":"2026-10-17T16:50:37.065Z","detectedPublishedAt":"2026-10-17T16:20:21.000Z","rights":"original"}],"claims":[{"id":"0e38906e-70d3-41c7-85f2-34d49f5e7e55","kind":"recommendation-relevant","statement":"Linux and Windows builds are listed for CUDA 12 and CUDA 13.","citations":[{"source":2,"supportProbability":0.63,"supports":false,"quote":null}]},{"id":"10a5aa35-34aa-4163-b328-fdfd22315353","kind":"capability","statement":"b11507 adds support for MoE cache over multiple GPUs.","citations":[{"source":2,"supportProbability":0.94,"supports":true,"quote":{"text":"llama: support MoE cache over multiple GPUs ( #30112 )","rights":"original"}}]},{"id":"857f92c0-871b-4f55-bd2f-2f58ce7a9fac","kind":"availability","statement":"The release lists macOS, iOS, Linux, Android, and Windows builds.","citations":[{"source":2,"supportProbability":0.77,"supports":true,"quote":null}]},{"id":"da054bba-d159-4bc6-ad71-83f83023dfc9","kind":"availability","statement":"b11507 is marked as a pre-release.","citations":[{"source":2,"supportProbability":0.95,"supports":true,"quote":null}]},{"id":"9fb81b32-df54-471a-b161-f77db0661fbf","kind":"capability","statement":"The release extends Gemma-4 GPU decode through OpenCL to E2B.","citations":[{"source":3,"supportProbability":0.93,"supports":true,"quote":{"text":"opencl: enable Gemma-4 E4B GPU decode","rights":"original"}}]},{"id":"80cb6d15-85b3-4f95-9b0e-3832df22dc30","kind":"capability","statement":"The release improves OpenCL support for FA and DK512 for Gemma-4.","citations":[{"source":3,"supportProbability":0.93,"supports":true,"quote":null}]},{"id":"5c439b5b-62b7-4fdd-a2a2-7ab66999e90b","kind":"deployment","statement":"The release lists builds for macOS, iOS, Linux, Android, Windows, and openEuler.","citations":[{"source":3,"supportProbability":0.8,"supports":true,"quote":null}]},{"id":"3c615db7-b200-4d74-bc80-70c4309552a8","kind":"capability","statement":"The release enables Gemma-4 E4B GPU decode through OpenCL.","citations":[{"source":3,"supportProbability":0.93,"supports":true,"quote":{"text":"opencl: enable Gemma-4 E4B GPU decode","rights":"original"}}]},{"id":"804c7cff-cb5e-407e-bbfd-625e3cb5d4c6","kind":"availability","statement":"The release is marked Pre-release.","citations":[{"source":3,"supportProbability":0.95,"supports":true,"quote":null}]},{"id":"48c33507-f770-4335-9df9-97b21d08f590","kind":"capability","statement":"The release adds MTP support for Mellum2.1.","citations":[{"source":4,"supportProbability":0.93,"supports":true,"quote":{"text":"model : MTP support for Mellum2.1 ( #30612 )","rights":"original"}}]},{"id":"a255f375-033b-4689-9c74-7e73184edfce","kind":"availability","statement":"b11695 is marked as a pre-release.","citations":[{"source":4,"supportProbability":0.95,"supports":true,"quote":null}]},{"id":"0b2ae5e2-c992-4969-b968-2441c17a7e44","kind":"availability","statement":"Builds are listed for macOS/iOS, Linux, Android, and Windows.","citations":[{"source":4,"supportProbability":0.74,"supports":true,"quote":null}]},{"id":"8b366cbf-90db-4b95-a88f-91fe57bd602c","kind":"capability","statement":"The release adds inline RoPE construction in Mellum.","citations":[{"source":4,"supportProbability":0.92,"supports":true,"quote":{"text":"model : inline rope construction in Mellum","rights":"original"}}]},{"id":"9b1f1dcf-31cb-4375-a791-f41f010798c7","kind":"capability","statement":"The release adds MTP support for Mellum.","citations":[{"source":4,"supportProbability":0.92,"supports":true,"quote":{"text":"model : MTP support for Mellum2.1 ( #30612 )","rights":"original"}}]},{"id":"f9ca9a3f-32b8-4b7b-a22a-002f3dd7aea4","kind":"capability","statement":"Paused downloads can be resumed, and incomplete downloads are remembered across restarts.","citations":[{"source":5,"supportProbability":0.75,"supports":true,"quote":null}]},{"id":"349e3c11-fdcc-4dfc-a292-6c4f3198b2c4","kind":"availability","statement":"b11714 is marked as a pre-release.","citations":[{"source":5,"supportProbability":0.86,"supports":true,"quote":null}]},{"id":"8dfe4670-e5f5-4d99-b6f4-79cc4d7a6b5c","kind":"capability","statement":"The Discover Models surface includes a curated catalog list with search and a details pane.","citations":[{"source":5,"supportProbability":0.84,"supports":true,"quote":{"text":"The Discover surface: the curated-catalog list with its search and","rights":"original"}}]},{"id":"8800fade-c639-4edc-9573-3f4324d31e6d","kind":"capability","statement":"The Discover interface includes download options and a standalone download progress bar.","citations":[{"source":5,"supportProbability":0.78,"supports":true,"quote":{"text":"download options, and the standalone download progress bar that replaces","rights":"original"}}]},{"id":"81b18e7b-84d0-455a-97de-cc6fa99feab8","kind":"capability","statement":"The Discover setting gates browsing and downloading Hugging Face GGUF models.","citations":[{"source":5,"supportProbability":0.8,"supports":true,"quote":{"text":"The flag gates browsing and downloading HuggingFace GGUF models, so it is","rights":"original"}}]},{"id":"e0ed379f-59d2-49a6-821b-afaab4c95431","kind":"capability","statement":"The new llama_batch_ext API supports mixed token and embedding batches and per-token state embeddings.","citations":[{"source":1,"supportProbability":0.98,"supports":true,"quote":{"text":"New llama_batch_ext extended batch API with llama_process() , supporting mixed token/embedding batches and per-token \"state\" embeddings for MTP and deepstack models #24669","rights":"original"}}]},{"id":"1f6053de-11c2-43e2-9534-26c05c2dc5b6","kind":"capability","statement":"The server adds a /v1/systemone API for decision models.","citations":[{"source":1,"supportProbability":0.98,"supports":true,"quote":{"text":"llama and server: new /v1/systemone API supporting five decision models - laya, julia-1, lev, openjev (+vision), kev #29818","rights":"original"}}]},{"id":"9ef57c20-6220-4cb2-b788-f0cc33242076","kind":"capability","statement":"The /v1/embeddings endpoint accepts typed vision, audio, and video content.","citations":[{"source":1,"supportProbability":0.98,"supports":true,"quote":{"text":"tools/server : new /v1/systemone endpoint for decision models #29818 and /v1/embeddings now accepts typed vision/audio/video content #29556","rights":"original"}}]},{"id":"1e82332a-d64b-4d1d-a286-5170982e5b9d","kind":"capability","statement":"The release adds support for the GLM-5.3-Flash 320B text-and-vision hybrid model and the Clef decision model with text and vision.","citations":[{"source":1,"supportProbability":0.98,"supports":true,"quote":{"text":"New models: GLM-5.3-Flash (GLM5-Next), a 320B text+vision hybrid model #27773 , and the Clef decision model, fully supported with both text and vision #29831 #29969","rights":"original"}}]},{"id":"a0a8aaf5-2da4-4fa6-8208-2c8450c923cc","kind":"availability","statement":"llama.cpp v0.6.0 was released on 05 Oct.","citations":[{"source":1,"supportProbability":0.9,"supports":true,"quote":null}]}]},"nextCursor":null,"license":{"catalog":{"name":"CC BY 4.0","spdx":"CC-BY-4.0","url":"https://creativecommons.org/licenses/by/4.0/","attribution":"wowsuchai","attributionUrl":"https://wowsuchai.withglint.app","covers":["catalog metadata","classifications","generated summaries","feed event metadata"]},"excluded":{"covers":["source quotations","linked source content"],"terms":"Not licensed by wowsuchai. Source quotations and linked source content keep their original rights."}}}