---

license: other

license_name: swift-open-license-1.0

license_link: https://huggingface.co/ukisai/Swift-Qwen3.8-Flash-Next/blob/main/LICENSE

library_name: gguf

pipeline_tag: image-text-to-text

gated: true

tags:

• gguf
• llama.cpp
• qwen3_8
• moe
• reasoning
• efficient-thinking
• token-efficient
• post-training
• agentic
• terminal-bench

datasets:

• ukisai/Qwen3.8-27B-multi-turn-agent-sft

base_model: ukisai/Swift-Qwen3.8-Flash-Next

base_model_relation: quantized

---

<div align="center">

<a href="https://ukisai.com"><img src="ukisai-banner.png" alt="UkisAI" style="width:100%;max-width:100%;height:auto;display:block;margin-bottom:0.6em;" /></a>

<div style="display:flex;justify-content:center;gap:0.6em;margin-bottom:1em;">

<a href="https://ukisai.com"><strong>Website</strong></a> &nbsp;&bull;&nbsp;

<a href="https://ukisai.com/products/swift"><strong>Learn more</strong></a> &nbsp;&bull;&nbsp;

<a href="https://huggingface.co/ukisai/Swift-1.5-Qwen3.8-Flash-Next-GGUF"><strong>GGUF</strong></a> &nbsp;&bull;&nbsp;

<a href="https://huggingface.co/ukisai/Swift-1.5-Qwen3.8-Flash-Next-GSQ-RCO-GGUF"><strong>GSQ-RCO GGUF</strong></a> &nbsp;&bull;&nbsp;

<a href="#evaluation"><strong>Evaluation</strong></a> &nbsp;&bull;&nbsp;

<a href="#license-and-access"><strong>Enterprise licensing</strong></a>

</div>

</div>

Swift 1.5 Qwen3.8-Flash-Next

GGUF quantizations. Derived directly from Swift 1.5 Qwen3.8-Flash-Next with llama.cpp. Run a chosen quantization tier with a current llama.cpp-compatible runtime such as llama-server.

Swift 1.5 Qwen3.8-Flash-Next is UkisAI's reasoning-efficient derivative of Qwen3.8-Flash-Next.

It uses 63.4% fewer thinking tokens, with a 1.8x speed up while keeping the accuracy loss <1% vs base on xhigh.

Demo

We gave base Qwen3.8-Flash-Next and Swift 1.5 Qwen3.8-Flash-Next the same prompt:

> Create a 3D endless runner that has the fast, playful feel of Subway Surfers, but make the world and characters your own. I want to run through a lively place, dodge things, collect rewards, and feel the pace build the longer I survive. Make it fun to control and visually memorable. Use your judgment for the setting, mechanics, and little details that make it feel like a real game. Build it so I can launch and play it locally, then tell me how to run it.

<video src="https://huggingface.co/ukisai/Swift-1.5-Qwen3.8-Flash-Next-GGUF/resolve/main/swift-1.5-flash-next-demo.mp4" controls autoplay muted loop playsinline style="width:100%;height:auto;border-radius:12px;"></video>

Try the game yourself here: https://ukisai.com/swift-games/flash-next

Base Qwen3.8-Flash-Next took 8 minutes 52 seconds to build its game. Swift 1.5 took 4 minutes 56 seconds.

Training approach

We made Swift Flash Next efficient by figuring out which tokens were linked to pathological overthinking and penalizing them without "attacking" the reasoning length directly then regained the accuracy with RL and OPD, leading to "compressed" token usage while maintaining accuracy.

Swift 1.5 produces shorter reasoning traces. In our testing, we also observe fewer overthinking errors.

This release also features our previously mentioned post-training methods adapted specifically for coding and long-horizon agent

work such as personal agents, terminal use and software engineering.

Our training data is viewable here: https://huggingface.co/datasets/ukisai/Qwen3.8-27B-multi-turn-agent-sft albeit it is not used out of the box, but rather re-sampled, turned into proper RL environments etc.

Evaluation

All scores compare the Qwen3.8-Flash-Next BF16 base with the Swift 1.5 BF16 checkpoint. Token columns report thinking tokens, except Terminal-Bench 2.1, which reports total generated output tokens.

<style>

.swift-table { width:100%; table-layout:fixed; border-collapse:separate; border-spacing:0; overflow:hidden; border:1px solid #27344A; border-radius:20px; background:#0D111B; font-family:-apple-system,BlinkMacSystemFont,"Segoe UI",Roboto,sans-serif; font-size:14px; color:#BFBDBD; }

.swift-table th { padding:13px 8px; text-align:center; font-weight:700; color:#AEB5C7; background:#0D111B; border-right:1px solid #27344A; border-bottom:1px solid #27344A; }

.swift-table td { padding:14px 8px; text-align:center; color:#BFBDBD; background:#0D111B; border-right:1px solid #27344A; border-bottom:1px solid #27344A; vertical-align:middle; overflow-wrap:break-word; }

.swift-table tr > :last-child { border-right:0; }

.swift-table tbody tr:last-child td { border-bottom:0; }

.swift-table .benchmark-heading { color:#B7BDCD; background:#0D111B; border-bottom:3px solid #7D45B5; }

.swift-table .score-heading { color:#F0C5FF; background:#52239E; border-bottom:3px solid #7D45B5; }

.swift-table .tokens-heading, .swift-table .median-heading { color:#D4E8FF; background:#304FC2; border-bottom:3px solid #5687E6; }

.swift-table .benchmark { padding-left:18px; text-align:left; color:#FFFFFF; font-weight:600; }

.swift-table .section { padding:12px 18px; text-align:left; color:#B489FF; background:#2A2541; font-weight:700; letter-spacing:.08em; text-transform:uppercase; border-top:1px solid #3A3159; border-bottom:1px solid #3A3159; }

.swift-table .swift { background:#171127; }

.swift-table thead tr:nth-child(2) .swift { color:#D3A0FF; }

.swift-table .reduction { color:#69BFFF; background:#101B2C; font-weight:700; }

.swift-table .detail { color:#8C94A8; font-size:12px; font-weight:500; }

@media (max-width: 640px) {

.swift-table { display:block !important; width:100% !important; max-width:100%; overflow-x:auto !important; -webkit-overflow-scrolling:touch; table-layout:auto !important; }

.swift-table th, .swift-table td { min-width:100px; }

.swift-table th:first-child, .swift-table td:first-child { min-width:160px; }

}

</style>

<table class="swift-table">

<thead>

<tr>

<th rowspan="2" class="benchmark-heading" style="width:32%;text-align:left;padding-left:18px;vertical-align:bottom;">Benchmark</th>

<th colspan="2" class="score-heading">Score</th>

<th colspan="3" class="tokens-heading">Mean tokens</th>

<th class="median-heading" style="width:14%;">Median tokens</th>

</tr>

<tr>

<th>Base</th><th class="swift">Swift 1.5</th>

<th>Base</th><th class="swift">Swift 1.5</th>

<th class="reduction">Reduction</th><th class="reduction">Reduction</th>

</tr>

</thead>

<tbody>

<tr><td class="section" colspan="7">General reasoning</td></tr>

<tr><td class="benchmark">GPQA-Diamond</td><td>89.80%</td><td class="swift">89.60%</td><td>17,683</td><td class="swift">7,823</td><td class="reduction">&darr; 55.8%</td><td class="reduction">&darr; 63.4%</td></tr>

<tr><td class="benchmark">MMLU-Pro</td><td>87.75%</td><td class="swift">87.20%</td><td>3,528</td><td class="swift">1,519</td><td class="reduction">&darr; 57.0%</td><td class="reduction">&darr; 24.0%</td></tr>

<tr><td class="benchmark">C-Eval</td><td>93.27%</td><td class="swift">93.60%</td><td>1,048</td><td class="swift">586</td><td class="reduction">&darr; 44.1%</td><td class="reduction">&darr; 7.1%</td></tr>

<tr><td class="benchmark">IFBench</td><td>73.20%</td><td class="swift">70.13%</td><td>8,310</td><td class="swift">4,411</td><td class="reduction">&darr; 46.9%</td><td class="reduction">&darr; 55.6%</td></tr>

<tr><td class="section" colspan="7">Mathematics</td></tr>

<tr><td class="benchmark">AIME 2026</td><td>98.67%</td><td class="swift">96.67%</td><td>23,015</td><td class="swift">15,806</td><td class="reduction">&darr; 31.3%</td><td class="reduction">&darr; 51.1%</td></tr>

<tr><td class="benchmark">HMMT (Nov 2025)</td><td>98.00%</td><td class="swift">97.33%</td><td>25,487</td><td class="swift">16,530</td><td class="reduction">&darr; 35.1%</td><td class="reduction">&darr; 54.7%</td></tr>

<tr><td class="section" colspan="7">Multimodal</td></tr>

<tr><td class="benchmark">ERQA</td><td>70.80%</td><td class="swift">69.30%</td><td>4,036</td><td class="swift">1,788</td><td class="reduction">&darr; 55.7%</td><td class="reduction">&darr; 47.7%</td></tr>

<tr><td class="section" colspan="7">Coding</td></tr>

<tr><td class="benchmark">LiveCodeBench v6</td><td>88.40%</td><td class="swift">90.39%</td><td>17,833</td><td class="swift">9,849</td><td class="reduction">&darr; 44.8%</td><td class="reduction">&darr; 52.0%</td></tr>

<tr><td class="section" colspan="7">Agentic coding</td></tr>

<tr><td class="benchmark">Terminal-Bench 2.1</td><td>67.64%</td><td class="swift">69.66%</td><td>40,591</td><td class="swift">45,428</td><td class="reduction">&uarr; 11.9%</td><td class="reduction">&darr; 17.9%</td></tr>

</tbody>

</table>

<details>

<summary><strong>How to reproduce</strong></summary>

<p style="font-size:13px;line-height:1.5;margin:8px 0;"><strong>Serving:</strong> BF16 · Qwen3 reasoning parser · context 262,144 · thinking xhigh · MTP disabled.<br>

<strong>Sampling:</strong> temperature 1.0 · top_p 0.95 · top_k 20 · min_p 0 · presence_penalty 0 · repetition_penalty 1.<br>

<strong>Benchmarks:</strong> five seeds (0–4) for the seeded question benchmarks; Terminal-Bench 2.1 uses five attempts per task. LiveCodeBench is full release v6 mean pass@1 over seeds.<br>

<strong>Terminal-Bench 2.1:</strong> Harbor 0.20.0 / Terminus-2 2.0.0, pinned 89-task dataset, JSON parser, interleaved thinking, temperature 1, top_p 1, 131,072-token server context, 3,600-second LLM call timeout and native per-task limits. Swift used concurrency 8; its context-recovery fix was applied during the run.</p>

<table style="display:table;width:100%;border-collapse:collapse;font-size:13px;line-height:1.3;margin:8px 0;">

<thead><tr><th style="padding:4px 8px;text-align:left;">Benchmark</th><th style="padding:4px 8px;text-align:right;">Output cap</th></tr></thead>

<tbody>

<tr><td style="padding:3px 8px;">GPQA-Diamond</td><td style="padding:3px 8px;text-align:right;">100,000</td></tr>

<tr><td style="padding:3px 8px;">MMLU-Pro</td><td style="padding:3px 8px;text-align:right;">100,000</td></tr>

<tr><td style="padding:3px 8px;">C-Eval</td><td style="padding:3px 8px;text-align:right;">16,384</td></tr>

<tr><td style="padding:3px 8px;">IFBench</td><td style="padding:3px 8px;text-align:right;">81,920</td></tr>

<tr><td style="padding:3px 8px;">AIME 2026</td><td style="padding:3px 8px;text-align:right;">250,000</td></tr>

<tr><td style="padding:3px 8px;">HMMT Nov 2025</td><td style="padding:3px 8px;text-align:right;">250,000</td></tr>

<tr><td style="padding:3px 8px;">ERQA</td><td style="padding:3px 8px;text-align:right;">100,000</td></tr>

<tr><td style="padding:3px 8px;">LiveCodeBench v6</td><td style="padding:3px 8px;text-align:right;">100,000</td></tr>

</tbody>

</table>

</details>

Efficiency across reasoning efforts

GPQA-Diamond at each reasoning_effort setting, Swift 1.5 against the base at the same setting:

<table class="swift-table" style="display:table;width:100%;table-layout:fixed;">

<colgroup><col style="width:16%"><col style="width:12%"><col style="width:12%"><col style="width:12%"><col style="width:12%"><col style="width:18%"><col style="width:18%"></colgroup>

<thead>

<tr>

<th rowspan="2" class="benchmark-heading" style="width:16%;text-align:left;padding-left:18px;vertical-align:bottom;">Reasoning effort</th>

<th colspan="2" class="score-heading" style="width:24%;">Score</th>

<th colspan="3" class="tokens-heading" style="width:42%;">Mean tokens</th>

<th class="median-heading" style="width:18%;">Median tokens</th>

</tr>

<tr>

<th>Base</th><th class="swift">Swift 1.5</th>

<th>Base</th><th class="swift">Swift 1.5</th>

<th class="reduction">Reduction</th><th class="reduction">Reduction</th>

</tr>

</thead>

<tbody>

<tr><td class="benchmark">Xhigh</td><td>89.80%</td><td class="swift">89.60%</td><td>17,683</td><td class="swift">7,823</td><td class="reduction">&darr; 55.8%</td><td class="reduction">&darr; 63.4%</td></tr>

<tr><td class="benchmark">Medium</td><td>86.36%</td><td class="swift">83.74%</td><td>4,157</td><td class="swift">2,483</td><td class="reduction">&darr; 40.3%</td><td class="reduction">&darr; 25.1%</td></tr>

<tr><td class="benchmark">Low</td><td>87.17%</td><td class="swift">84.75%</td><td>3,966</td><td class="swift">2,645</td><td class="reduction">&darr; 33.3%</td><td class="reduction">&darr; 19.7%</td></tr>

</tbody>

</table>

At xhigh, Swift 1.5 trails base by 0.20 percentage points while using 55.8% fewer mean and 63.4% fewer median thinking tokens. Medium and low save tokens but also lose 2.62 and 2.42 percentage points respectively.

Quantized models

| Format | Repository | Runtime |

| --- | --- | --- |

| AWQ INT4 (W4A16) | Swift-1.5-Qwen3.8-Flash-Next-W4A16-AWQ | vLLM (compressed-tensors) |

| AutoRound INT4 (W4A16) | Swift-1.5-Qwen3.8-Flash-Next-W4A16-AutoRound | vLLM (auto-round) |

| NVFP4 | Swift-1.5-Qwen3.8-Flash-Next-NVFP4 | NVIDIA Blackwell |

| GGUF | Swift-1.5-Qwen3.8-Flash-Next-GGUF | llama.cpp |

| GSQ-RCO GGUF (compact 2–3 bit) | Swift-1.5-Qwen3.8-Flash-Next-GSQ-RCO-GGUF | llama.cpp |

GGUF quantizations

<table class="swift-table" style="display:table;width:100%;table-layout:fixed;">

<thead><tr>

<th class="benchmark-heading" style="width:20%;text-align:left;padding-left:18px;white-space:normal;">File</th>

<th class="score-heading" style="width:12%;white-space:normal;">Size</th>

<th class="score-heading" style="width:17%;white-space:normal;">KLD wikitext @512</th>

<th class="tokens-heading" style="width:17%;white-space:normal;">KLD wikitext @32k</th>

<th class="tokens-heading" style="width:17%;white-space:normal;">99% KLD @32k</th>

<th class="median-heading" style="width:17%;white-space:normal;">Top-p @32k</th>

</tr></thead>

<tbody>

<tr><td class="benchmark">Q8_0</td><td>188.3 GB</td><td class="swift">0.0297</td><td class="reduction">0.0220</td><td class="reduction">0.230</td><td>94.91%</td></tr>

<tr><td class="benchmark">Q6_K</td><td>168.1 GB</td><td>0.0324</td><td class="reduction">0.0252</td><td class="reduction">0.256</td><td>94.34%</td></tr>

<tr><td class="benchmark">Q5_K_L</td><td>151.1 GB</td><td>0.0788</td><td class="reduction">0.0575</td><td class="reduction">0.610</td><td>91.61%</td></tr>

<tr><td class="benchmark">Q4_K_L</td><td>139.3 GB</td><td>0.1051</td><td class="reduction">0.0740</td><td class="reduction">0.791</td><td>90.14%</td></tr>

<tr><td class="benchmark">Q5_K_M</td><td>134.7 GB</td><td>0.1641</td><td class="reduction">0.1386</td><td class="reduction">1.654</td><td>87.34%</td></tr>

<tr><td class="benchmark">Q5_K_S</td><td>128.1 GB</td><td>0.1777</td><td class="reduction">0.1482</td><td class="reduction">1.763</td><td>86.92%</td></tr>

<tr><td class="benchmark">Q4_K_M</td><td>119.6 GB</td><td>0.1512</td><td class="reduction">0.1100</td><td class="reduction">1.254</td><td>88.20%</td></tr>

<tr><td class="benchmark">Q3_K_XL</td><td>119.3 GB</td><td>0.3718</td><td class="reduction">0.3250</td><td class="reduction">3.303</td><td>79.79%</td></tr>

<tr><td class="benchmark">Q4_K_S</td><td>113.1 GB</td><td>0.1779</td><td class="reduction">0.1311</td><td class="reduction">1.420</td><td>87.01%</td></tr>

<tr><td class="benchmark">Q4_1</td><td>111.3 GB</td><td>0.2532</td><td class="reduction">0.1812</td><td class="reduction">2.048</td><td>85.07%</td></tr>

<tr><td class="benchmark">Q2_K_L</td><td>107.1 GB</td><td>0.5425</td><td class="reduction">0.3867</td><td class="reduction">3.654</td><td>77.28%</td></tr>

<tr><td class="benchmark">Q4_0</td><td>100.6 GB</td><td>0.2524</td><td class="reduction">0.2593</td><td class="reduction">2.798</td><td>82.53%</td></tr>

<tr><td class="benchmark">IQ4_NL</td><td>100.3 GB</td><td>0.1261</td><td class="reduction">0.0926</td><td class="reduction">1.010</td><td>88.96%</td></tr>

<tr><td class="benchmark">IQ4_XS</td><td>97.7 GB</td><td>0.1305</td><td class="reduction">0.0929</td><td class="reduction">1.010</td><td>88.91%</td></tr>

<tr><td class="benchmark">IQ3_M</td><td>93.2 GB</td><td>0.2005</td><td class="reduction">0.1482</td><td class="reduction">1.605</td><td>86.22%</td></tr>

<tr><td class="benchmark">Q3_K_L</td><td>93.2 GB</td><td>0.3842</td><td class="reduction">0.3300</td><td class="reduction">3.259</td><td>79.44%</td></tr>

<tr><td class="benchmark">Q3_K_M</td><td>92.0 GB</td><td>0.3210</td><td class="reduction">0.2260</td><td class="reduction">2.401</td><td>82.73%</td></tr>

<tr><td class="benchmark">IQ3_XS</td><td>91.9 GB</td><td>0.2211</td><td class="reduction">0.1644</td><td class="reduction">1.738</td><td>85.26%</td></tr>

<tr><td class="benchmark">Q3_K_S</td><td>89.4 GB</td><td>0.3253</td><td class="reduction">0.2284</td><td class="reduction">2.399</td><td>82.59%</td></tr>

<tr><td class="benchmark">IQ3_XXS</td><td>88.0 GB</td><td>0.2868</td><td class="reduction">0.2185</td><td class="reduction">2.350</td><td>83.30%</td></tr>

<tr><td class="benchmark">Q2_K</td><td>80.9 GB</td><td>0.5703</td><td class="reduction">0.4091</td><td class="reduction">3.832</td><td>76.66%</td></tr>

<tr><td class="benchmark">IQ2_M</td><td>80.4 GB</td><td>0.3769</td><td class="reduction">0.2759</td><td class="reduction">2.708</td><td>80.76%</td></tr>

<tr><td class="benchmark">IQ2_S</td><td>77.8 GB</td><td>0.4846</td><td class="reduction">0.4331</td><td class="reduction">4.196</td><td>76.47%</td></tr>

<tr><td class="benchmark">IQ2_XS</td><td>77.7 GB</td><td>0.5136</td><td class="reduction">0.4033</td><td class="reduction">3.798</td><td>76.93%</td></tr>

<tr><td class="benchmark">IQ2_XXS</td><td>75.2 GB</td><td>0.6052</td><td class="reduction">0.4472</td><td class="reduction">4.002</td><td>75.58%</td></tr>

<tr><td class="benchmark">IQ1_M</td><td>72.0 GB</td><td>0.9067</td><td class="reduction">0.6928</td><td class="reduction">5.392</td><td>69.57%</td></tr>

<tr><td class="benchmark">IQ1_S</td><td>70.1 GB</td><td>1.1601</td><td class="reduction">0.9343</td><td class="reduction">6.726</td><td>65.15%</td></tr>

</tbody>

</table>

Mean KL divergence against the Swift 1.5 BF16 source, lower is better. wikitext @512 is wikitext-2

test, 100 windows of 512 tokens. wikitext @32k is wikitext-2 test, 9 windows of 32,768 tokens, scoring

the second half of each window, so every scored token sees at least 16k tokens of context.

99% KLD is the 99th-percentile divergence on the same 32k run. Top-p is top-token agreement with BF16

on the 32k run.

Each tier is in its own folder, split into several parts. llama.cpp loads every part from the first one:

llama-server -hf ukisai/Swift-1.5-Qwen3.8-Flash-Next-GGUF:Q4_K_M
# or, after downloading a folder:
llama-server -m Q4_K_M/Swift-1.5-Qwen3.8-Flash-Next-Q4_K_M-00001-of-00003.gguf

<table class="swift-table" style="display:table;width:100%;table-layout:fixed;">

<thead><tr>

<th class="benchmark-heading" style="width:55%;text-align:left;padding-left:18px;white-space:normal;">Use case</th>

<th class="score-heading" style="width:45%;white-space:normal;">Pick</th>

</tr></thead>

<tbody>

<tr><td class="benchmark">128 GB unified memory, everyday use</td><td class="swift">IQ4_XS</td></tr>

<tr><td class="benchmark">Long agentic runs, strict tool-call formatting</td><td class="swift">Q6_K or higher</td></tr>

<tr><td class="benchmark">Maximum fidelity</td><td class="swift">Q8_0</td></tr>

</tbody>

</table>

<details>

<summary><strong>Recipe</strong></summary>

All 27 tiers were built with llama.cpp commit d2e5458 from a BF16 conversion of the published Swift 1.5

safetensors. They use bartowski's exact per-tensor type map read from

bartowski/Qwen3.8-Flash-Next-GGUF, with an

importance matrix computed on this model from his calibration-v6 corpus (-c 512, 583 chunks,

--parse-special). Every file was checked against BF16 on the harness above.

</details>

License and access

Swift 1.5 Qwen3.8-Flash-Next is a derivative of Qwen3.8-Flash-Next

(Copyright (c) 2026 Qwen, Qwen Community License 1.0).

UkisAI's contribution, including the adapted weights, is licensed under the

Swift Open License v1.0.

See NOTICE for the change notice and

attribution details.

Personal, research, educational, evaluation, and commercial use of the Swift contribution are free for individuals

and organizations with gross annual revenue, including affiliates, of up to US$1,000,000. Above that threshold,

commercial use requires a separate Swift Enterprise License. Contact UkisAI for terms.

The base model's terms still apply to it. Under the Qwen Community License 1.0, organizations that run a

Model-as-a-Service or AI Work Assistant business need a separate license from Qwen before any commercial use,

and products above 100 million monthly active users or US$20 million monthly revenue must prominently display

the model name. Nothing in the Swift Open License limits your rights in Qwen3.8-Flash-Next itself under the

Qwen Community License.

Citation

~~~bibtex

@misc{swift-qwen3.8-flash-next,

title = {Swift 1.5 Qwen3.8-Flash-Next},

author = {UkisAI},

year = {2026},

url = {https://huggingface.co/ukisai/Swift-Qwen3.8-Flash-Next}

}

~~~

Acknowledgements

We acknowledge the NVIDIA Innovation Lab,

Amazon Web Services, and

Google Cloud for providing compute credits and

infrastructure support for Swift's development, training, and evaluation.