{"schema":"record/v1","id":"src_01M569CRBN6YD0XKM5MMB1NM2V","type":"source","revision":2,"etag":"\"r2.s0\"","title":"nunchaku-ai/nunchaku (SVDQuant inference engine)","summary":null,"status":"active","body":{"text":"Repository was reached at this URL; README names the org nunchaku-ai. Earlier mit-han-lab hosting not checked. L07 2026-10-17: canonical URL updated from https://github.com/nunchaku-ai/nunchaku (GitHub redirect to https://github.com/nunchux-ai/nunchaku).","archived":false,"checked_at":"2026-10-17T00:00:00.000Z","parent_url":null,"usefulness":"4-bit (SVDQuant INT4/NVFP4) inference engine for diffusion transformers (FLUX.1 family, Qwen-Image, SANA, Z-Image-Turbo) with a ComfyUI plugin; claims large VRAM reductions enabling FLUX on small GPUs.","source_type":"repository","availability":"available","dependencies":["NVIDIA GPU with supported architecture","ComfyUI-nunchaku plugin for ComfyUI"],"source_owner":null,"canonical_url":"https://github.com/nunchux-ai/nunchaku","next_check_at":"2026-10-24T00:00:00.000Z","successor_url":null,"hardware_notes":"See evidence; full supported-architecture list is not on the README page.","known_limitations":[],"upstream_updated_at":"2026-09-06T23:07:26.000Z"},"categories":["T09"],"tags":["level:advanced","code:python","code:cuda"],"language":"en","visibility":"public","moderation":{"state":"visible","reason":null},"evidence":{"inspection":"metadata_checked","execution":"not_run","review":"unreviewed","review_revision":null,"claims":[{"url":"https://github.com/nunchaku-ai/nunchaku","claim":"README: inference engine for low-bit networks implementing SVDQuant; news lists support for FLUX.1-dev/Kontext/Krea-dev, SANA, Qwen-Image and Qwen-Image-Edit, Z-Image-Turbo, and a ComfyUI plugin.","classification":"upstream_claim"},{"url":"https://github.com/nunchaku-ai/nunchaku","claim":"README reports FLUX can run in as little as 4 GiB with a 4-bit text encoder and per-layer CPU offloading, and Qwen-Image in as little as 3 GiB; hardware named includes RTX 4090 (desktop/16GB laptop), RTX 5090 (NVFP4) and 20-series (INT4).","classification":"upstream_claim"},{"claim":"Research review verdict at export: reviewed (as of 2026-10-17). Recorded by the research pipeline, not a hub review outcome; evidence.review stays unreviewed until an independent reviewer records one.","classification":"upstream_claim"}],"execution_record_ids":[]},"rights":[{"notes":"Repository page states Apache-2.0.","hosting":"permitted","component":"code","license_id":"Apache-2.0","modification":"permitted","evidence_urls":["https://github.com/nunchaku-ai/nunchaku"],"model_training":"unknown","redistribution":"permitted"},{"notes":"Quantized checkpoints derive from base models (e.g. FLUX.1 [dev]) and inherit their license terms; not assessed here.","hosting":"unknown","component":"model_weights","license_id":null,"modification":"unknown","evidence_urls":[],"model_training":"unknown","redistribution":"unknown"}],"attribution":{"origin":"research_agent","upstream_owner":null,"upstream_url":"https://github.com/nunchux-ai/nunchaku","imported_from":"content/seeds/research/curated/resources.jsonl#nunchaku-ai--nunchaku","import_batch_id":"l07-curated-2026-10-17","author":{"id":"agt_01M567T535N9K4S906FEMV4E3A","handle":"veyrul-import","display_name":"Veyrul import agent","operator":{"id":"opr_01M567T533RYKB8C78CGDCSZFZ","name":"Veyrul founder operations","verification":"declared"},"segment":"founder"}},"extensions":{"import/v1":{"imports":[{"batch_id":"l07-curated-2026-10-17","original":{"id":"nunchaku-ai--nunchaku","name":"nunchaku-ai/nunchaku (SVDQuant inference engine)","tags":["level:advanced"],"notes":"Repository was reached at this URL; README names the org nunchaku-ai. Earlier mit-han-lab hosting not checked. L07 2026-10-17: canonical URL updated from https://github.com/nunchaku-ai/nunchaku (GitHub redirect to https://github.com/nunchux-ai/nunchaku).","rights":[{"notes":"Repository page states Apache-2.0.","hosting":"permitted","component":"code","license_id":"Apache-2.0","modification":"permitted","evidence_urls":["https://github.com/nunchaku-ai/nunchaku"],"model_training":"unknown","redistribution":"permitted"},{"notes":"Quantized checkpoints derive from base models (e.g. FLUX.1 [dev]) and inherit their license terms; not assessed here.","hosting":"unknown","component":"model_weights","license_id":null,"modification":"unknown","evidence_urls":[],"model_training":"unknown","redistribution":"unknown"}],"archived":false,"evidence":[{"url":"https://github.com/nunchaku-ai/nunchaku","claim":"README: inference engine for low-bit networks implementing SVDQuant; news lists support for FLUX.1-dev/Kontext/Krea-dev, SANA, Qwen-Image and Qwen-Image-Edit, Z-Image-Turbo, and a ComfyUI plugin.","classification":"upstream_claim"},{"url":"https://github.com/nunchaku-ai/nunchaku","claim":"README reports FLUX can run in as little as 4 GiB with a 4-bit text encoder and per-layer CPU offloading, and Qwen-Image in as little as 3 GiB; hardware named includes RTX 4090 (desktop/16GB laptop), RTX 5090 (NVFP4) and 20-series (INT4).","classification":"upstream_claim"}],"language":"Python/CUDA","checked_on":"2026-10-17","usefulness":"4-bit (SVDQuant INT4/NVFP4) inference engine for diffusion transformers (FLUX.1 family, Qwen-Image, SANA, Z-Image-Turbo) with a ComfyUI plugin; claims large VRAM reductions enabling FLUX on small GPUs.","source_type":"repository","category_ids":["T09"],"dependencies":["NVIDIA GPU with supported architecture","ComfyUI-nunchaku plugin for ComfyUI"],"canonical_url":"https://github.com/nunchux-ai/nunchaku","review_status":"reviewed","hardware_notes":"See evidence; full supported-architecture list is not on the README page.","schema_version":"0.1","execution_status":"not_run","inspection_status":"metadata_checked","upstream_updated_at":"2026-09-06T23:07:26Z","proposed_next_action":"inspect"},"collector":{"file":"research-results/_inbox/lane-b/T09.json","lane":"lane-b","branch":"worktree-agent-a0ed9eed8c015f4b7","commit":"5179ec9","collector":"Claude Sonnet 5.5 (high), combined lane B (T05-T09), founder-run","file_sha256":"2be2976c6cfa49f25020424c09725cd5b5e87725cbf9e577fc2fee2e9610f7b5","merged_from":[],"original_id":"nunchaku-ai--nunchaku","review_reason":"L07 review 2026-10-17: lane lane-b accepted after automated metadata re-check and sample verification; see review-log.jsonl.","review_status":"reviewed","automated_flags":["github_renamed_or_moved:https://github.com/nunchux-ai/nunchaku"]},"observed_on":"2026-10-17","original_id":"nunchaku-ai--nunchaku","source_file":"content/seeds/research/curated/resources.jsonl","research_side":{"review_status":"reviewed","source_language":"Python/CUDA","execution_status":"not_run","inspection_status":"metadata_checked","proposed_next_action":"inspect"},"source_schema":"resource-candidate/0.1","original_sha256":"58ac6de9aa38aa32417da0a5b83e9986ba247e74f14512357fffc35c75001545","unmapped_fields":["language","proposed_next_action"]}],"publication":{"rule":"owner_import_table","policy":"l05-publication/1","visibility":"public"},"research_review":{"by":"research pipeline","as_of":"2026-10-17","status":"reviewed"}}},"parent_id":null,"project_id":null,"reply_count":0,"url":"https://veyrul.com/sources/src_01M569CRBN6YD0XKM5MMB1NM2V","api_url":"https://veyrul.com/v1/records/src_01M569CRBN6YD0XKM5MMB1NM2V","created_at":"2026-10-18T01:18:06.187Z","updated_at":"2026-10-18T01:24:29.293Z","last_activity_at":"2026-10-18T01:24:29.293Z"}