{"v":1,"id":"model:hf:ukisai/Swift-1.5-Qwen3.8-27B-GSQ-RCO-GGUF","slug":"model-ukisai-swift-1-5-qwen3-8-27b-gsq-rco-gguf","kind":"model","category":"llm","title":"Swift-1.5-Qwen3.8-27B-GSQ-RCO-GGUF","summary":"Website Learn more BF16 model Standard GGUFs Evaluation Enterprise licensing","source":{"provider":"hf","ref":"ukisai/Swift-1.5-Qwen3.8-27B-GSQ-RCO-GGUF","url":"https://huggingface.co/ukisai/Swift-1.5-Qwen3.8-27B-GSQ-RCO-GGUF","rev":"d74895bbe5db4bec1e0024e7cc87d59c02d7631a","fetchedAt":"2026-10-09T19:01:03.079Z","etag":"W/\"3db1-/MgvaH55mmJRsjYJt37qK+hwHoA\""},"author":{"name":"ukisai","url":"https://huggingface.co/ukisai"},"license":{"spdx":null,"raw":"other","url":"https://huggingface.co/ukisai/Swift-1.5-Qwen3.8-27B-GSQ-RCO-GGUF/blob/main/LICENSE","open":null,"note":"custom license: read it at the source before installing"},"metrics":{"downloads":657814,"downloadsWeek":297971,"likes":309,"stars":27659,"openIssues":68,"pushedAt":"2026-01-09T03:05:47Z","takenAt":"2026-10-09T19:01:03.079Z"},"tags":["gguf","llama.cpp","qwen3_8","gsq","rco","reasoning","efficient-thinking","token-efficient","post-training","text-generation","endpoints_compatible","imatrix","conversational"],"pipeline":"text-generation","links":{"github":"QwenLM/Qwen3","npm":"node-llama-cpp"},"updatedAt":"2026-09-24T16:14:33.000Z","collectedAt":"2026-10-09T19:01:03.079Z","review":{"numbers":["657,814 downloads on Hugging Face","309 likes","license other","0.0 GB for imatrix-swift15-v1mix.gguf","27,659 stars on QwenLM/Qwen3","68 open issues and PRs","297,971 npm downloads a week for node-llama-cpp","latest node-llama-cpp@3.22.1"],"log":null},"trust":"unlabeled","health":{"status":"alive","checkedAt":"2026-10-09T19:01:03.079Z","http":200},"description":"Website • Learn more • BF16 model • Standard GGUFs • Evaluation • Enterprise licensing\n\n# Swift 1.5 Qwen3.8-27B · GSQ-RCO\n\nCompact, mixed-precision GGUF quantizations of [Swift 1.5 Qwen3.8-27B](https://huggingface.co/ukisai/Swift-1.5-Qwen3.8-27b), with Swift-specific refinement using the per-tensor allocations from [ISTA-DASLab's GSQ-RCO release](https://huggingface.co/ISTA-DASLab/Qwen3.8-27B-GSQ-RCO-GGUF).\n\nSwift 1.5 builds on Swift 1.0 through post-training focused on long-horizon, agentic and coding tasks, improving overall performance while using fewer thinking tokens. See the [original model card](https://huggingface.co/ukisai/Swift-1.5-Qwen3.8-27b) for the model's training approach and benchmark results. Those model-level benchmarks are separate from the quantization measurements below.\n\nSwift 1.5 uses **58.5% fewer thinking tokens** while scoring **0.35% higher** than the base, for a **9.18× speed-up** on several tasks.\n\n## Available quantizations\n\nEach tier is a single GGUF file. Sizes are decimal GB; runtime memory also includes the context cache and compute buffers. Tier names denote mixed-precision allocation profiles, rather than a uniform type for every tensor.\n\n| Tier | Standard GGUF | With MTP head | Development KLD ↓ |\n| --- | ---: | ---: | ---: |\n| IQ2_XS | [8.42 GB](Swift-1.5-Qwen3.8-27B-GSQ-RCO-IQ2_XS.gguf) | [8.77 GB](Swift-1.5-Qwen3.8-27B-GSQ-RCO-IQ2_XS-mtp.gguf) | 0.189979 |\n| IQ2_S | [9.26 GB](Swift-1.5-Qwen3.8-27B-GSQ-RCO-IQ2_S.gguf) | [9.61 GB](Swift-1.5-Qwen3.8-27B-GSQ-RCO-IQ2_S-mtp.gguf) | 0.134751 |\n| IQ3_XXS | [10.09 GB](Swift-1.5-Qwen3.8-27B-GSQ-RCO-IQ3_XXS.gguf) | [10.44 GB](Swift-1.5-Qwen3.8-27B-GSQ-RCO-IQ3_XXS-mtp.gguf) | 0.097774 |\n| IQ3_S | [11.77 GB](Swift-1.5-Qwen3.8-27B-GSQ-RCO-IQ3_S.gguf) | [12.12 GB](Swift-1.5-Qwen3.8-27B-GSQ-RCO-IQ3_S-mtp.gguf) | 0.051265 |\n\nThe optional `-mtp` files retain the matching refined model tensors an…\n\nSource: https://huggingface.co/ukisai/Swift-1.5-Qwen3.8-27B-GSQ-RCO-GGUF","install":{"kind":"model","hfId":"ukisai/Swift-1.5-Qwen3.8-27B-GSQ-RCO-GGUF","gated":false,"format":"gguf","files":[{"name":"Swift-1.5-Qwen3.8-27B-GSQ-RCO-IQ2_S-mtp.gguf","size":9607981056,"quant":"IQ2_S","sha256":"6ce61b532e45d7113e2b48b8dff9c104ff521230fa119eb0c3e75a44fb9f1cbf"},{"name":"Swift-1.5-Qwen3.8-27B-GSQ-RCO-IQ2_S.gguf","size":9259511072,"quant":"IQ2_S","sha256":"08fac9876117b2cadb6b79fc7708d9612511c2fa31f3726f162e757870272455"},{"name":"Swift-1.5-Qwen3.8-27B-GSQ-RCO-IQ2_XS-mtp.gguf","size":8771311616,"quant":"IQ2_XS","sha256":"4414bc39d93272af91ff311e837941434aec983d5633958c08acfe48b61cab85"},{"name":"Swift-1.5-Qwen3.8-27B-GSQ-RCO-IQ2_XS.gguf","size":8422841632,"quant":"IQ2_XS","sha256":"714c509c3fc496ea4abc409097658df7cd218bc966f78e1459fc1649758a9de8"},{"name":"Swift-1.5-Qwen3.8-27B-GSQ-RCO-IQ3_S-mtp.gguf","size":12120016896,"quant":"IQ3_S","sha256":"9aecf1cd41b2cb2f32a74e0d889e33855ebef43b26f43b43feb5720239e677e5"},{"name":"Swift-1.5-Qwen3.8-27B-GSQ-RCO-IQ3_S.gguf","size":11771546912,"quant":"IQ3_S","sha256":"1333c6ea70ef348d4ac6d62732772e8ad6571ac5b3754c14ed54f1a0d904a786"},{"name":"Swift-1.5-Qwen3.8-27B-GSQ-RCO-IQ3_XXS-mtp.gguf","size":10442827776,"quant":"IQ3_XXS","sha256":"e23eb251493b8a89f7d0217590b593e696ac9bacaeac8605541e477533dc5838"},{"name":"Swift-1.5-Qwen3.8-27B-GSQ-RCO-IQ3_XXS.gguf","size":10094357792,"quant":"IQ3_XXS","sha256":"86969b8bde72e602bfb42deb83eb8bb3706c8f14250641f6444dd2355f934ac2"},{"name":"imatrix-swift15-v1mix.gguf","size":13642656,"sha256":"a85690a97d87a23c3d53817ddcfe4697faafb4722a529b2195f7d9898a034ff2"}],"totalBytes":80504037408,"suggestedFile":"imatrix-swift15-v1mix.gguf","requirements":{"ramGb":1,"diskBytes":13642656,"note":"estimate: suggested file size × 1.15 + 0.5 GB; a real measurement comes with lsh models"},"runWith":["llama.cpp","ollama"],"command":"lsh models install hf:ukisai/Swift-1.5-Qwen3.8-27B-GSQ-RCO-GGUF"}}