KRAFTON/Raon-1-mini-preview
text-generation model by KRAFTON
From the publisher
Model card & documentation
Source previewRead the publisher’s intended use, setup instructions, evaluations and limitations. The original model card is the source of truth.
Raon-1 mini is a 30B-A4B Mixture-of-Experts (MoE) language model, designed and trained from scratch by KRAFTON AI. It demonstrates strong capabilities in Korean conversation and agentic tasks. As part of the Raon open model family, it is released under the Apache 2.0 license. Raon-1 mini uses a hybrid Mixture-of-Experts (MoE) architecture, consisting of 18 linear attention layers and 7 full attention layers. CategoryBenchmarkMoEDense Raon-1 mini preview(30B-A4B)Qwen3.6-35B-A3BGemma 4 26B-A4BQwen3.8-27BMuse Glimmer 30B KnowledgeMMLU Pro76.585.982.985.784.0 AA-Omniscience3.5−7.2−46.6−0.5−34.7 ReasoningGPQA Diamond84.383.280.789.782.6 AIME 202693.392.987.198.892.9 HMMT Feb 202684.184.579.995.874.6 Instruction and multi-turnIFBench84.266.680.982.478.7 MultiChallenge59.058.258.463.663.4 CodingLiveCodeBench v681.681.283.479.377.4 SciCode38.735.639.644.643.6 SWE-Verified¹70.863.261.674.672.0 Terminal-Bench Lite47.063.352.067.070.7 AgenticGDPVal²66.067.457.563.567.3 AutomationBench³29.724.936.258.732.1 BrowseComp Plus55.734.832.258.052.8 BFCL v4 Overall59.864.765.872.451.8 Long contextAA-LCR67.369.367.080.381.3 GDP.pdf⁴5.86.63.611.28.4 Korean conversationChatbot Korean (internal)⁵92.974.590.690.187.0 Judge and user simulator: gpt-5.6-luna for all evaluations. Red = Raon-1 mini preview is best within the MoE group (Raon-1 mini preview, Qwen3.6-35B-A3B, Gemma 4 26B-A4B). Dense models (Qwen3.8-27B, Muse Glimmer 30B) are shown for reference only. ¹ SWE-Verified: evaluated with the mini-SWE-agent harness. ² GDPVal…
Read the full model card ↗ · Preview checked 2026-10-31T15:32:23.915Z
Inside the original model card — Document outline
- Raon-1 mini
- Highlights
- Model Overview
- Benchmark Results
- Sampling parameters
- Speculative Decoding
- Quickstart
- RTX 5090 (32 GB)
- DGX Spark (GB10)
- Agentic Use
- Hermes Agent
- Pi
Headings are captured from the source. Links open the publisher’s document, not a locally hosted copy.
Links from the model card
References supplied by the publisher, not independently verified endorsements. Check the destination before downloading files or entering credentials.
Documentation belongs to its respective authors. Reported project/model license: apache-2.0. A listing is not a grant of reuse or training rights. Confirm the document’s own terms at the source.
Model overview
KRAFTON/Raon-1-mini-preview is a text-generation model repository published by KRAFTON on Hugging Face. The source reports the transformers library.
This page summarizes Hub metadata. For intended use, training data, evaluation results and limitations, consult the original model card.
Model facts
- Task
- text-generation
- Library
- transformers
- Recent downloads (30 days)
- 3,280
- Cumulative likes
- 228
- Hugging Face trending score
- 64
- Reported safetensors parameters
- 28,783,218,402
- Architecture
- Raon1MiniForCausalLM
- License
- apache-2.0
- Access
- Not gated by Hugging Face
- Created
- 2026-10-21T00:14:48.000Z
- Last modified
- 2026-10-27T08:07:23.000Z
Compatibility and lineage
Language coverage was not reported.
Base-model lineage was not included.
Parameter count is not a RAM/VRAM requirement. Check precision, quantization, context length and runtime compatibility in the model card; no hardware or API-cost claim is inferred here.
Use and evaluate
Check license terms and access requirements first. Review the model card’s documented loading instructions, evaluations and safety limitations. Benchmark scores are not imported by this directory, and popularity is not an accuracy ranking.
No model weights are downloaded or executed by AltAPIs. Never enable remote model code without reviewing it.
Source and freshness
Source: Hugging Face Hub. Metadata observed 2026-10-31T06:49:10.134Z. Daily imports are snapshots, not real-time monitoring.
Popularity and source listings do not establish security, suitability, licensing rights or benchmark performance.