Update README.md
Browse files
README.md
CHANGED
|
@@ -6,215 +6,8 @@ sdk: static
|
|
| 6 |
pinned: true
|
| 7 |
---
|
| 8 |
|
| 9 |
-
|
| 10 |
|
| 11 |
-
|
| 12 |
|
| 13 |
-
|
| 14 |
-
|
| 15 |
-
### Thailand's Leading Sovereign AI Company
|
| 16 |
-
|
| 17 |
-
[](https://iapp.co.th)
|
| 18 |
-
[](https://iapp.co.th/docs)
|
| 19 |
-
[](https://discord.gg/kYcpmdEcS2)
|
| 20 |
-
|
| 21 |
-
**Empowering businesses with enterprise-grade AI solutions built for Thai language and local use cases.**
|
| 22 |
-
|
| 23 |
-
[Explore Our APIs](https://iapp.co.th/docs) | [View Pricing](https://iapp.co.th/pricing) | [Contact Us](https://iapp.co.th/contact-us)
|
| 24 |
-
|
| 25 |
-
</div>
|
| 26 |
-
|
| 27 |
-
---
|
| 28 |
-
|
| 29 |
-
## About Us
|
| 30 |
-
|
| 31 |
-
**iApp Technology Co., Ltd.** is Thailand's leading sovereign AI company, providing enterprise-grade artificial intelligence solutions built specifically for Thai language and local use cases. We serve **100+ enterprise customers** including banks, insurance companies, and government agencies, with over **50 million API calls** processed.
|
| 32 |
-
|
| 33 |
-
- **Founded:** With a mission to make AI accessible to Thai enterprises
|
| 34 |
-
- **Certifications:** ISO 29110, iBeta PAD Level 1 & 2
|
| 35 |
-
- **Compliance:** PDPA and GDPR compliant
|
| 36 |
-
- **Headquarters:** Bangkok, Thailand
|
| 37 |
-
|
| 38 |
-
Learn more at [iapp.co.th/about-us](https://iapp.co.th/about-us)
|
| 39 |
-
|
| 40 |
-
---
|
| 41 |
-
|
| 42 |
-
## Featured Models
|
| 43 |
-
|
| 44 |
-
### Chinda Thai LLM
|
| 45 |
-
|
| 46 |
-
Our flagship Thai language model, optimized for Thai text understanding and generation.
|
| 47 |
-
|
| 48 |
-
| Model | Parameters | Description | Link |
|
| 49 |
-
|-------|------------|-------------|------|
|
| 50 |
-
| **[chinda-qwen3-4b](https://huggingface.co/iapp/chinda-qwen3-4b)** | 4B | Full-precision Thai LLM for text generation | [Try API](https://iapp.co.th/docs/llm/chinda-4b) |
|
| 51 |
-
| **[chinda-qwen3-4b-gguf](https://huggingface.co/iapp/chinda-qwen3-4b-gguf)** | 4B | GGUF format for local deployment (Ollama, LM Studio) | [Tutorial](https://iapp.co.th/openmodels/chinda-llm-ollama-tutorial) |
|
| 52 |
-
|
| 53 |
-
**FREE API Access:** Chinda Thai LLM 4B API is FREE until Dec 31, 2025!
|
| 54 |
-
[Get Started](https://iapp.co.th/blog/introducing-chinda-llm-api-free-thai-language-model)
|
| 55 |
-
|
| 56 |
-
---
|
| 57 |
-
|
| 58 |
-
## Open Datasets
|
| 59 |
-
|
| 60 |
-
We believe in advancing AI research through open collaboration. Here are our publicly available datasets:
|
| 61 |
-
|
| 62 |
-
### Evaluation & Benchmarking
|
| 63 |
-
|
| 64 |
-
| Dataset | Size | Downloads | Description |
|
| 65 |
-
|---------|------|-----------|-------------|
|
| 66 |
-
| [OpenThaiEval](https://huggingface.co/datasets/iapp/openthaieval) | 1,200 | 52 | Comprehensive evaluation dataset for Thai LLMs |
|
| 67 |
-
| [OpenAI HumanEval-TH](https://huggingface.co/datasets/iapp/openai_humaneval-th) | 164 | 164 | Thai translation of OpenAI's HumanEval |
|
| 68 |
-
| [Code Generation Lite-TH](https://huggingface.co/datasets/iapp/code_generation_lite-th) | 111 | 111 | Code generation evaluation in Thai context |
|
| 69 |
-
|
| 70 |
-
### Mathematics & Reasoning
|
| 71 |
-
|
| 72 |
-
| Dataset | Size | Downloads | Description |
|
| 73 |
-
|---------|------|-----------|-------------|
|
| 74 |
-
| [AIME 2024-TH](https://huggingface.co/datasets/iapp/aime_2024-th) | 30 | 30 | Thai AIME 2024 math problems |
|
| 75 |
-
| [Math-500-TH](https://huggingface.co/datasets/iapp/math-500-th) | 500 | 500 | Thai mathematics reasoning dataset |
|
| 76 |
-
| [AIMO Validation AIME-TH](https://huggingface.co/datasets/iapp/aimo-validation-aime-th) | 90 | 90 | AI Mathematical Olympiad validation |
|
| 77 |
-
|
| 78 |
-
### Training & Specialized Datasets
|
| 79 |
-
|
| 80 |
-
| Dataset | Size | Downloads | Description |
|
| 81 |
-
|---------|------|-----------|-------------|
|
| 82 |
-
| [Thai-R1-Distill-SFT](https://huggingface.co/datasets/iapp/Thai-R1-Distill-SFT) | 10K | 10K+ | SFT dataset for reasoning capabilities |
|
| 83 |
-
| [Thai Handwriting Dataset](https://huggingface.co/datasets/iapp/thai_handwriting_dataset) | 13.6K | 13.6K+ | Thai handwritten text for OCR |
|
| 84 |
-
| [RAG Thai Laws](https://huggingface.co/datasets/iapp/rag_thai_laws) | 42.8K | 42.8K+ | Thai legal documents for RAG systems |
|
| 85 |
-
| [Thai Famous People Images](https://huggingface.co/datasets/iapp/thai_famous_people_images_dataset) | 4.2K | 4.2K+ | Thai celebrity face dataset |
|
| 86 |
-
|
| 87 |
-
Browse all datasets: [huggingface.co/datasets/iapp](https://huggingface.co/datasets/iapp)
|
| 88 |
-
|
| 89 |
-
---
|
| 90 |
-
|
| 91 |
-
## Research & Publications
|
| 92 |
-
|
| 93 |
-
Our research team actively contributes to advancing AI for Thai language:
|
| 94 |
-
|
| 95 |
-
### 2025
|
| 96 |
-
|
| 97 |
-
- **[OpenThaiGPT 1.6 and R1](https://arxiv.org/abs/2504.01789)** - Thai-Centric Open Source and Reasoning Large Language Models
|
| 98 |
-
*Sumeth Yuenyong, Thodsaporn Chay-intr, Kobkrit Viriyayudhakorn*
|
| 99 |
-
|
| 100 |
-
- **[LLaVAC: Fine-tuning LLaVA as a Multimodal Sentiment Classifier](https://arxiv.org/abs/2502.02938)**
|
| 101 |
-
*T. Chay-intr, Y. Chen, K. Viriyayudhakorn, T. Theeramunkong*
|
| 102 |
-
|
| 103 |
-
### 2024
|
| 104 |
-
|
| 105 |
-
- **[OpenThaiGPT 1.5: Technical Report](https://link.springer.com/chapter/10.1007/978-3-031-97144-0_19)** - NLDB 2024
|
| 106 |
-
*Sumeth Yuenyong, Kobkrit Viriyayudhakorn, et al.*
|
| 107 |
-
|
| 108 |
-
View all publications: [iapp.co.th/researches/publications](https://iapp.co.th/researches/publications)
|
| 109 |
-
|
| 110 |
-
---
|
| 111 |
-
|
| 112 |
-
## Enterprise AI APIs
|
| 113 |
-
|
| 114 |
-
Beyond open source, we provide production-ready AI APIs:
|
| 115 |
-
|
| 116 |
-
| Category | APIs | Documentation |
|
| 117 |
-
|----------|------|---------------|
|
| 118 |
-
| **E-KYC** | Face Verification, Liveness Detection, ID Card OCR | [Docs](https://iapp.co.th/docs/category/-electronic-know-your-customer-e-kyc) |
|
| 119 |
-
| **LLM** | Chinda Thai LLM, DeepSeek, Thanoy Legal AI | [Docs](https://iapp.co.th/docs/category/-large-language-model) |
|
| 120 |
-
| **OCR** | Thai ID Card, Passport, Driver License, Documents | [Docs](https://iapp.co.th/docs/category/-thai-document-optical-character-recognition-thai-document-ocr) |
|
| 121 |
-
| **Speech** | Thai ASR (Speech-to-Text), TTS (Text-to-Speech) | [Docs](https://iapp.co.th/docs/category/%EF%B8%8F-speech-technology) |
|
| 122 |
-
| **NLP** | Translation, Sentiment Analysis, Summarization | [Docs](https://iapp.co.th/docs/category/-thai-natural-language-processing) |
|
| 123 |
-
| **Smart City** | License Plate Recognition, Object Detection | [Docs](https://iapp.co.th/docs/category/%EF%B8%8F-smart-city-ai) |
|
| 124 |
-
|
| 125 |
-
[View Full API Catalog](https://iapp.co.th/docs) | [Pricing](https://iapp.co.th/pricing)
|
| 126 |
-
|
| 127 |
-
---
|
| 128 |
-
|
| 129 |
-
## Quick Start
|
| 130 |
-
|
| 131 |
-
### Using Chinda via API
|
| 132 |
-
|
| 133 |
-
```python
|
| 134 |
-
from openai import OpenAI
|
| 135 |
-
|
| 136 |
-
client = OpenAI(
|
| 137 |
-
api_key="your-iapp-api-key",
|
| 138 |
-
base_url="https://api.iapp.co.th/v3/llm/chinda-4b"
|
| 139 |
-
)
|
| 140 |
-
|
| 141 |
-
response = client.chat.completions.create(
|
| 142 |
-
model="chinda-4b",
|
| 143 |
-
messages=[
|
| 144 |
-
{"role": "user", "content": "สวัสดีครับ ช่วยแนะนำอาหารไทยให้หน่อย"}
|
| 145 |
-
]
|
| 146 |
-
)
|
| 147 |
-
|
| 148 |
-
print(response.choices[0].message.content)
|
| 149 |
-
```
|
| 150 |
-
|
| 151 |
-
### Using Chinda Locally (Ollama)
|
| 152 |
-
|
| 153 |
-
```bash
|
| 154 |
-
# Pull the model
|
| 155 |
-
ollama pull hf.co/iapp/chinda-qwen3-4b-gguf
|
| 156 |
-
|
| 157 |
-
# Run chat
|
| 158 |
-
ollama run hf.co/iapp/chinda-qwen3-4b-gguf
|
| 159 |
-
```
|
| 160 |
-
|
| 161 |
-
[Full Ollama Tutorial](https://iapp.co.th/openmodels/chinda-llm-ollama-tutorial) | [LM Studio Tutorial](https://iapp.co.th/openmodels/chinda-lm-studio-tutorial)
|
| 162 |
-
|
| 163 |
-
---
|
| 164 |
-
|
| 165 |
-
## Community & Support
|
| 166 |
-
|
| 167 |
-
- **Discord:** [Join our community](https://discord.gg/kYcpmdEcS2)
|
| 168 |
-
- **GitHub:** [github.com/iapp-technology](https://github.com/iapp-technology)
|
| 169 |
-
- **Documentation:** [iapp.co.th/docs](https://iapp.co.th/docs)
|
| 170 |
-
- **Blog:** [iapp.co.th/blog](https://iapp.co.th/blog)
|
| 171 |
-
- **API Status:** [status.iapp.co.th](https://status.iapp.co.th/status/iapp-ai-apis)
|
| 172 |
-
|
| 173 |
-
---
|
| 174 |
-
|
| 175 |
-
## Citation
|
| 176 |
-
|
| 177 |
-
If you use our models or datasets in your research, please cite:
|
| 178 |
-
|
| 179 |
-
```bibtex
|
| 180 |
-
@misc{iapp_chinda_2024,
|
| 181 |
-
author = {iApp Technology Research Team},
|
| 182 |
-
title = {Chinda: Thai Large Language Model},
|
| 183 |
-
year = {2024},
|
| 184 |
-
publisher = {iApp Technology},
|
| 185 |
-
url = {https://iapp.co.th/openmodels/chinda-opensource-llm}
|
| 186 |
-
}
|
| 187 |
-
|
| 188 |
-
@misc{iapp_datasets_2024,
|
| 189 |
-
author = {iApp Technology Research Team},
|
| 190 |
-
title = {iApp Open Datasets for Thai AI Research},
|
| 191 |
-
year = {2024},
|
| 192 |
-
publisher = {iApp Technology},
|
| 193 |
-
url = {https://iapp.co.th/researches/datasets}
|
| 194 |
-
}
|
| 195 |
-
```
|
| 196 |
-
|
| 197 |
-
---
|
| 198 |
-
|
| 199 |
-
## Contact
|
| 200 |
-
|
| 201 |
-
**iApp Technology Co., Ltd.**
|
| 202 |
-
|
| 203 |
-
- **Website:** [https://iapp.co.th](https://iapp.co.th)
|
| 204 |
-
- **Email:** sale@iapp.co.th
|
| 205 |
-
- **Phone:** 086-322-5858
|
| 206 |
-
- **Location:** Bangkok, Thailand
|
| 207 |
-
|
| 208 |
-
[Contact Us](https://iapp.co.th/contact-us) | [Careers](https://iapp.co.th/careers)
|
| 209 |
-
|
| 210 |
-
---
|
| 211 |
-
|
| 212 |
-
<div align="center">
|
| 213 |
-
|
| 214 |
-
**Made with in Thailand**
|
| 215 |
-
|
| 216 |
-
*Advancing Thai AI for the World*
|
| 217 |
-
|
| 218 |
-
[](https://iapp.co.th)
|
| 219 |
-
|
| 220 |
-
</div>
|
|
|
|
| 6 |
pinned: true
|
| 7 |
---
|
| 8 |
|
| 9 |
+
**Thailand's Leading Sovereign AI Company** - Enterprise-grade AI APIs for Thai language: LLM, E-KYC, OCR, Speech & NLP. Trusted by 100+ companies with 50M+ API calls. ISO 29110 & iBeta certified.
|
| 10 |
|
| 11 |
+
**Open Models:** Chinda Thai LLM 4B (FREE API until Dec 2025!) | **Datasets:** 12+ Thai AI datasets
|
| 12 |
|
| 13 |
+
[**Visit iapp.co.th**](https://iapp.co.th) | [**API Docs**](https://iapp.co.th/docs) | [**Pricing**](https://iapp.co.th/pricing) | [**Open Models**](https://iapp.co.th/openmodels/chinda-opensource-llm) | [**Datasets**](https://iapp.co.th/researches/datasets) | [**Publications**](https://iapp.co.th/researches/publications) | [**Discord**](https://discord.gg/kYcpmdEcS2)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|