# openthai2.0-qwen3.8-27b by iApp Technology: Open Model Source: https://savrn.com/models/openthai2-0-qwen3-8-27b Markdown alternate of the page above; the site index is https://savrn.com/llms.txt --- ## Runs On What it takes to serve openthai2.0-qwen3.8-27b (27.8B parameters): the memory its weights need at each precision, and the cheapest way to rent enough data-center GPUs to hold them. | Precision | Weights | Memory needed | Cheapest setup | Per hour | Also fits | | --- | --- | --- | --- | --- | --- | | 16-bit | 55.6 GB | 66.7 GB | 1x [MI300X](https://savrn.com/ai-index/pricing/gpus/mi300x) (192 GB) Vultr | $1.85 | [1x H100](https://savrn.com/ai-index/pricing/gpus/h100) $1.99 · [1x MI325X](https://savrn.com/ai-index/pricing/gpus/mi325x) $2.00 | | 8-bit | 27.8 GB | 33.3 GB | 1x [MI300X](https://savrn.com/ai-index/pricing/gpus/mi300x) (192 GB) Vultr | $1.85 | [1x H100](https://savrn.com/ai-index/pricing/gpus/h100) $1.99 · [1x MI325X](https://savrn.com/ai-index/pricing/gpus/mi325x) $2.00 | | 4-bit | 13.9 GB | 16.7 GB | 1x [MI300X](https://savrn.com/ai-index/pricing/gpus/mi300x) (192 GB) Vultr | $1.85 | [1x H100](https://savrn.com/ai-index/pricing/gpus/h100) $1.99 · [1x MI325X](https://savrn.com/ai-index/pricing/gpus/mi325x) $2.00 | Memory is the weights at that precision plus 20% for the runtime and a short context; a long context needs more. Prices are the lowest on-demand hourly rates in the [SAVRN Index](https://savrn.com/ai-index/pricing/gpus), read Oct 7, 2026. [openthai2.0-qwen3.8-27b on every accelerator the SAVRN Index prices, at every precision](https://savrn.com/models/openthai2-0-qwen3-8-27b/gpus) ## Model Card By iApp Technology, published under apache-2.0, revision c48ac175714d. ### OpenThai2.0 - Opensource Thai Knowledge, Document, and Agentic AI Release notes - v2.0.3 (1 Sep 2026) — Thai factual recall repaired: the freerecall training data carried a phantom-citation register ("บทความระบุ...") and single-shape questions; cleaned and re-drilled with model-written paraphrases. Paraphrase battery 85/104 (from 79), Thai knowledge battery 21/21, OCR-Eval-104 CER 0.084 (from 0.093), HumanEval 0.982, IFEval-TH 0.837. Known residual: some numeric facts still scatter at temperature 0.7 — use low temperature for factual queries. - v2.0.2 (31 Aug 2026) — identifies as OpenThai 2.0 (iApp Technology + AIEAT) and adds a Thai-language safety guardrail; stock chat template. Printed-OCR CER about +0.02 vs v2.0.1, other benchmarks unchanged. - v2.0.1 (30 Aug 2026) — restores closed-book Thai knowledge lost in v2.0.0 (thanks to Dr. Panutat Tejasen's ThaiEval-2026 v3). - v2.0.0 — initial release. Earlier builds: revision tags v2.0.1, v2.0.0. Per-version benchmark tables: CHANGELOGS.md. โมเดล AI โอเพนซอร์ส ด้านความรู้ภาษาไทย เอกสารไทย และงานเอเจนต์ · เปิดตัวโดย iApp Technology ร่วมกับ สมาคมผู้ประกอบการปัญญาประดิษฐ์ประเทศไทย (AIEAT) · Apache 2.0 [Read the full model card (2,913 words)](https://savrn.com/models/openthai2-0-qwen3-8-27b/card) ## Configuration Architecture Qwen3_5ForConditionalGeneration Context length (tokens) 262,144 Layers 64 Hidden size 5,120 Feed-forward size 17,408 Attention heads 24 Key/value heads 4 Head dimension 256 Vocabulary size 248,320 Model type qwen3_5 ## Identity and Version Repository iapp/openthai2.0-qwen3.8-27b Publisher iApp Technology Task Image and text to text Modality Image and text Library Not stated by the source Parameters 27.8B parameters Languages th, en Revision c48ac175714d024605de02938974fd0d36c91eaa First published 2026-08-19 Last updated 2026-10-04 ## Files and Weights 49 files, 56.5 GB in total. The weights are 19 files totalling 56.5 GB in bin, pt, pth, safetensors. Weights19 files · 56.5 GB Configuration12 files · 202.7 KB Tokenizer2 files · 20.0 MB Documentation3 files · 41.4 KB Other12 files · 1.3 MB Repository1 file · 2.7 KB Every file | File | Type | Size | SHA-256 | | --- | --- | --- | --- | | adapter/adapter_model.safetensors | Weights | 934.0 MB | 25988c4cc05a | | adapter/rng_state_0.pth | Weights | 15.2 KB | 45559f439153 | | adapter/rng_state_1.pth | Weights | 15.2 KB | 8f7ded10956b | | adapter/rng_state_2.pth | Weights | 15.2 KB | 592fc1f0141b | | adapter/scheduler.pt | Weights | 1.5 KB | 85e1be670d09 | | adapter/training_args.bin | Weights | 8.7 KB | d7a6d33f1d06 | | model-00000.safetensors | Weights | 5.1 GB | c4449ad8b9a3 | | model-00001.safetensors | Weights | 4.6 GB | 1cbe1fb78c5d | | model-00002.safetensors | Weights | 4.5 GB | 634cce002008 | | model-00003.safetensors | Weights | 4.6 GB | ac796ef97abd | | model-00004.safetensors | Weights | 4.5 GB | 0c8d6537bf0c | | model-00005.safetensors | Weights | 4.6 GB | 203404fb6e81 | | model-00006.safetensors | Weights | 4.6 GB | b980c91ff706 | | model-00007.safetensors | Weights | 4.6 GB | d2982b83ddd2 | | model-00008.safetensors | Weights | 4.5 GB | d4ee02cbb2b9 | | model-00009.safetensors | Weights | 4.6 GB | 8b59d10a2d90 | | model-00010.safetensors | Weights | 4.6 GB | 832ce7f3fc61 | | model-00011.safetensors | Weights | 4.6 GB | d57a3722e4e6 | | model-00012.safetensors | Weights | 209.8 MB | 65f310d7ad56 | | adapter/adapter_config.json | Configuration | 1.1 KB | — | | adapter/additional_config.json | Configuration | 67 B | — | | adapter/args.json | Configuration | 20.2 KB | — | | adapter/trainer_state.json | Configuration | 21.3 KB | — | | adapter/zero_to_fp32.py | Configuration | 34.9 KB | — | | args.json | Configuration | 20.2 KB | — | | config.json | Configuration | 3.7 KB | — | | generation_config.json | Configuration | 214 B | — | | model.safetensors.index.json | Configuration | 99.0 KB | — | | preprocessor_config.json | Configuration | 390 B | — | | processor_config.json | Configuration | 1.2 KB | — | | video_preprocessor_config.json | Configuration | 385 B | — | | CHANGELOGS.md | Documentation | 5.7 KB | — | | README.md | Documentation | 30.5 KB | — | | adapter/README.md | Documentation | 5.3 KB | — | | adapter/latest | Other | 14 B | — | | assets/banner.png | Other | 210.9 KB | 62e4c609cc70 | | assets/bfcl.png | Other | 94.4 KB | — | | assets/cases/docbench_parliament_year_60_00000903_p0181.jpg | Other | 100.4 KB | 88b52c2de4d9 | | assets/cases/hw_cpeoph_20.jpg | Other | 3.5 KB | — | | assets/cases/hw_disjoint_48.jpg | Other | 2.8 KB | — | | assets/cases/wikisource_l4_ws4_53969.jpg | Other | 199.4 KB | bb2369c94fb2 | | assets/scoreboard.png | Other | 232.4 KB | a65c681692f6 | | assets/siamai-logo.png | Other | 45.1 KB | — | | assets/social/hero-en.webp | Other | 175.5 KB | 9d30ebad1e16 | | assets/social/hero-th.webp | Other | 193.8 KB | feaf7e906ef1 | | chat_template.jinja | Other | 9.0 KB | — | | .gitattributes | Repository | 2.7 KB | — | | tokenizer.json | Tokenizer | 20.0 MB | 06b9509352d2 | | tokenizer_config.json | Tokenizer | 1.2 KB | — | ## License and Download License apache-2.0 Access Open weights, no gate Download size 56.5 GB [Download from iApp Technology](https://huggingface.co/iapp/openthai2.0-qwen3.8-27b) Released by iApp Technology through its official repository on Hugging Face. [Read the license](https://www.apache.org/licenses/LICENSE-2.0). ## Built From - Derived from [Qwen/Qwen3.8-27B](https://savrn.com/models/qwen3-8-27b) ## Memory Requirements | Precision | Weights in memory | | --- | --- | | As published | 56.5 GB | | 16-bit | 55.6 GB | | 8-bit | 27.8 GB | | 4-bit | 13.9 GB | Weights only, from the published parameter count; the key-value cache and runtime add to this. ## Built on This Model - Quantized from[openthai2.0-qwen3.8-27b-MLX-4bit](https://savrn.com/models/openthai2-0-qwen3-8-27b-mlx-4bit) - Derived from[openthai2.0-qwen3.8-27b-MLX-4bit](https://savrn.com/models/openthai2-0-qwen3-8-27b-mlx-4bit) ## Questions About openthai2.0-qwen3.8-27b ### How much GPU memory does openthai2.0-qwen3.8-27b need? About 66.7 GB at 16-bit and 16.7 GB at 4-bit: the weights (27.8B parameters) plus a working margin. A long context needs more. ### What is the cheapest GPU to run openthai2.0-qwen3.8-27b on? At 16-bit, 1x MI300X from $1.85 an hour; at 4-bit, 1x MI300X from $1.85 an hour, at the lowest on-demand prices the SAVRN Index lists. ### Can I use openthai2.0-qwen3.8-27b commercially? Yes. openthai2.0-qwen3.8-27b is released under Apache License 2.0. The Apache License 2.0 is a permissive open-source license. It permits commercial use, modification and redistribution. It requires keeping the license and copyright notices and any NOTICE file, stating significant changes, and it includes an express patent grant from contributors. ### What is openthai2.0-qwen3.8-27b's context length? 262,144 tokens, from the maximum position embeddings in its published configuration. ## Similar Models Model · Image and text to text ### [Qwen3.8-27B](https://savrn.com/models/qwen3-8-27b) [Qwen](https://savrn.com/model-publishers/qwen) Following the widespread community adoption of the Qwen3.5 and Qwen3.6 series, we are pleased to introduce Qwen3.8, the most capable generation in the Qwen open-model family to date. Built on the architectural foundation of Qwen3.5, Qwen3.8 delivers substantial gains across coding, professional work, research, and long-horizon agentic tasks. Qwen3.8-27B brings these advances to a compact, deployment-friendly dense model: a native vision-language model that understands images and videos, with flexible thinking control, designed to carry complex, multi-step tasks through to completion with greater reliability. For streamlined integration, we recommend using Qwen3.8 via APIs. Qwen3.8 can be… Open weights apache-2.0 27.8B parameters 262,144 tokens transformers [View model](https://savrn.com/models/qwen3-8-27b) Model · Image and text to text ### [Qwen3.8-27B-FP8](https://savrn.com/models/qwen3-8-27b-fp8) [Qwen](https://savrn.com/model-publishers/qwen) Following the widespread community adoption of the Qwen3.5 and Qwen3.6 series, we are pleased to introduce Qwen3.8, the most capable generation in the Qwen open-model family to date. Built on the architectural foundation of Qwen3.5, Qwen3.8 delivers substantial gains across coding, professional work, research, and long-horizon agentic tasks. Qwen3.8-27B brings these advances to a compact, deployment-friendly dense model: a native vision-language model that understands images and videos, with flexible thinking control, designed to carry complex, multi-step tasks through to completion with greater reliability. For streamlined integration, we recommend using Qwen3.8 via APIs. Qwen3.8 can be… Open weights apache-2.0 27.8B parameters 262,144 tokens transformers [View model](https://savrn.com/models/qwen3-8-27b-fp8) Model · Image and text to text ### [Qwen3.6-27B](https://savrn.com/models/qwen3-6-27b) [Qwen](https://savrn.com/model-publishers/qwen) Following the February release of the Qwen3.5 series, we're pleased to share the first open-weight variant of Qwen3.6. Built on direct feedback from the community, Qwen3.6 prioritizes stability and real-world utility, offering developers a more intuitive, responsive, and genuinely productive coding experience. This release delivers substantial upgrades, particularly in For more details, please refer to our blog post Qwen3.6-27B. Empty cells (--) indicate scores not yet available or not applicable. For streamlined integration, we recommend using Qwen3.6 via APIs. Below is a guide to use Qwen3.6 via OpenAI-compatible API. Qwen3.6 can be served via APIs with popular inference frameworks. In… Open weights apache-2.0 27.8B parameters 262,144 tokens transformers [View model](https://savrn.com/models/qwen3-6-27b) Model · Image and text to text ### [Qwen3.5-27B](https://savrn.com/models/qwen3-5-27b) [Qwen](https://savrn.com/model-publishers/qwen) Over recent months, we have intensified our focus on developing foundation models that deliver exceptional utility and performance. Qwen3.5 represents a significant leap forward, integrating breakthroughs in multimodal learning, architectural efficiency, reinforcement learning scale, and global accessibility to empower developers and enterprises with unprecedented capability and efficiency. For more details, please refer to our blog post Qwen3.5. WMT24++: a harder subset of WMT24 after difficulty labeling and rebalancing; we report the averaged scores on 55 languages using XCOMET-XXL. Empty cells (--) indicate scores not yet available or not applicable. Empty cells (--) indicate scores not… Open weights apache-2.0 27.8B parameters 262,144 tokens transformers [View model](https://savrn.com/models/qwen3-5-27b) Model · Image and text to text ### [JEV-27B-VL](https://savrn.com/models/jev-27b-vl) [AutoTrust AI Lab](https://savrn.com/model-publishers/autotrust) autotrust/JEV-27B-VL is autotrust/JEV-27B with vision. Every step below is one System 1 decision: a camera image or a screenshot in, a probability for every action out, in a single forward pass. Robot arm: pick and place from a camera image. At every step System 1 looks at the top camera image and answers two questions: is the target left or right of the gripper, and above or below it? The arm moves accordingly and halves its step whenever an answer flips. It grasps the cube, carries it and drops it in the tray (MuJoCo simulation). 75% of 20 random scenes completed, the best of the JEV family. Every cube it grasped ended in the tray (15 of 15); every miss was a grasp 3.0–3.7 cm off target.… Open weights apache-2.0 27.8B parameters 262,144 tokens transformers [View model](https://savrn.com/models/jev-27b-vl) Model · Image and text to text ### [Qwen3.8-27B-AWQ-INT4](https://savrn.com/models/qwen3-8-27b-awq-int4) [Cyankiwi](https://savrn.com/model-publishers/cyankiwi) Following the widespread community adoption of the Qwen3.5 and Qwen3.6 series, we are pleased to introduce Qwen3.8, the most capable generation in the Qwen open-model family to date. Built on the architectural foundation of Qwen3.5, Qwen3.8 delivers substantial gains across coding, professional work, research, and long-horizon agentic tasks. Qwen3.8-27B brings these advances to a compact, deployment-friendly dense model: a native vision-language model that understands images and videos, with flexible thinking control, designed to carry complex, multi-step tasks through to completion with greater reliability. For streamlined integration, we recommend using Qwen3.8 via APIs. Qwen3.8 can be… Open weights apache-2.0 27.8B parameters 262,144 tokens transformers [View model](https://savrn.com/models/qwen3-8-27b-awq-int4) ## iApp Technology [All models and datasets](https://savrn.com/model-publishers/iapp) ## Versions - [c48ac175714d](https://savrn.com/models/openthai2-0-qwen3-8-27b/versions/c48ac175714d) · current 2026-10-04 ## Explore More - [All image and text to text models](https://savrn.com/models/tasks/image-and-text-to-text) - [All models under apache-2.0](https://savrn.com/models/licenses/apache-2-0) - [Model comparisons](https://savrn.com/models/comparisons) - [The model directory](https://savrn.com/models) - [Open model prices by host](https://savrn.com/ai-index/pricing/open-models) ## Source - Repository metadata, read 2026-10-04. - [Hugging Face record](https://huggingface.co/iapp/openthai2.0-qwen3.8-27b) - [How the hub is built](https://savrn.com/model-hub/methodology)