{"version":1,"type":"story","url":"https://digestai.news/story/mtmd-add-nemotron-3-5-transcribe-support-by-pskrunner14-pull-request-3","json":"https://digestai.news/story/mtmd-add-nemotron-3-5-transcribe-support-by-pskrunner14-pull-request-3.json","markdown":"https://digestai.news/story/mtmd-add-nemotron-3-5-transcribe-support-by-pskrunner14-pull-request-3.md","slug":"mtmd-add-nemotron-3-5-transcribe-support-by-pskrunner14-pull-request-3","headline":"Mtmd : add Nemotron 3.5 Transcribe support by pskrunner14 · Pull Request #31471 · ggml-org/llama.cpp","summary":"The update, submitted by a contributor named `pskrunner14`, integrates the model into the project’s multimodal toolkit, enabling audio-to-text conversion capabilities alongside existing video/image processing features.\n\nThe change resolves a conflict with the main branch, with the exception of a small delta in quantization work for IQ2\\(NL\\) and IQ3\\(NL\\) optimizations. The pull request, labeled as superseding prior work, was approved by **CISCappro** and merged into the master branch. The project’s documentation notes the inclusion of hidden Unicode characters, which may affect code interpretation. No new features or benchmarks were explicitly detailed in the post, but the addition expands Llama.cpp’s multimodal functionality.","keyPoints":["Llama.cpp now supports **Nemotron 3.5 Transcribe** for audio transcription via a merged pull request","Conflict resolved except for minor quantization optimizations for IQ2\\(NL\\)/IQ3\\(NL\\)","Update approved by **CISCappro** and merged into the project’s master branch"],"whyItMatters":"It may also attract users seeking open-source audio processing tools without vendor lock-in.","category":{"slug":"hardware","name":"Hardware & Compute","url":"https://digestai.news/category/hardware"},"entities":{"companies":[],"models":["Nemotron 3.5 Transcribe"],"people":["pskrunner14","CISCappro"]},"firstPublishedAt":"2026-10-31T00:00:00Z","updatedAt":"2026-10-31T00:00:00Z","sourceCount":1,"hasPrimarySource":true,"sources":[{"outlet":"github.com","title":"mtmd : add Nemotron 3.5 Transcribe support by pskrunner14 · Pull Request #31471 · ggml-org/llama.cpp","url":"https://github.com/ggml-org/llama.cpp/pull/31471","publishedAt":"2026-10-31T00:00:00Z","type":"primary","primary":true,"lead":true}],"sourceNotes":null,"discussions":[{"site":"Reddit","url":"https://www.reddit.com/r/LocalLLaMA/comments/1xmu0iz/mtmd_add_nemotron_35_transcribe_support_by/","points":null}],"thread":null,"cite":{"text":"Digest AI, \"Mtmd : add Nemotron 3.5 Transcribe support by pskrunner14 · Pull Request #31471 · ggml-org/llama.cpp\", 31 October 2026, https://digestai.news/story/mtmd-add-nemotron-3-5-transcribe-support-by-pskrunner14-pull-request-3","publisher":"Digest AI","title":"Mtmd : add Nemotron 3.5 Transcribe support by pskrunner14 · Pull Request #31471 · ggml-org/llama.cpp","datePublished":"2026-10-31T00:00:00Z","url":"https://digestai.news/story/mtmd-add-nemotron-3-5-transcribe-support-by-pskrunner14-pull-request-3"},"generatedBy":"Written by Digest AI's editorial model from the linked sources; the sources are the record.","license":"Headlines, digests and key points are written by Digest AI and may be quoted with a link to the story page. Linked articles belong to their publishers. Terms: https://digestai.news/terms#reuse"}