AFS (AI Foundry Service)
All-in-one No-Code Inference and Fine-tuning Solutions. Providing FFM and Open-source models to Empower AI Applications.
AFS (AI Foundry Service)
Taiwan AI Cloud introduces enterprise-grade GenAI solutions based on Formosa Foundation Model (FFM), including AFS ModelSpace (Model-as-a-Service) and AFS Platform + AFS Cloud (Integrated LLM Fine-tuning and Deployment Services). AFS supports enhanced Traditional Chinese models such as Llama3.1-FFM and FFM-Mixtral-8x7B, as well as various open-source models like Meta-Llama3.1 and the Taiwan-localized TAIDE model. Whether for inference or fine-tuning services, AFS provides the ideal solution. With low entry costs, it helps you save on massive AIHPC expenses and quickly advance into AI 2.0 application!
AFS (AI Foundry Service)：One-stop Solution
AFS ModelSpace
Model-as-a-Service
Directly use various FFM and open-source models with dual-mode pricing options for you to choose from.
✓ Easy to operate: Quickly deploy and start mainstream large language model AI services in the cloud.
✓ Cost transparency: Pay-as-you-go pricing. Simplify the pricing method to evaluate and control costs more easily.
✓ Cloud hosting: Automation tools and management services that simplify the administration tasks of large language models.
✓ Dual-mode billing: Billing based on either token usage or model deployment time, depending on preference.
| AFS ModelSpace | Public Mode Charged by token amount | Private Mode Charged by model deployment time |
|---|---|---|
| Formosa Foundation Models (FFM) | ⭐Llama3.2-FFM ⭐Llama3.1-FFM Llama3-FFM FFM-Mistral ⭐FFM-embedding-v2 FFM-embedding | Llama3.1-FFM Llama3-FFM FFM-Mistral FFM-Llama2-v2 FFM-Llama2 FFM-embedding-v2 FFM-embedding |
| Open-Source Models | ⭐TAIDE ⭐Meta-Llama3.3 Meta-Llama3.1 Qwen2.5-72B | ⭐TAIDE ⭐Meta-Llama3.3 Meta-Llama3.1 Meta-Llama2 Qwen2.5-72B Meta-CodeLlama |
*Please refer to AFS ModelSpace for complete products and prices.
AFS Platform
AFS Cloud
Integrated LLM Fine-tuning and Deployment Services
Provide enterprises with an one-stop solution to quickly build their own large language models from training to deployment.
✓ Reduce Training Costs: Provide pre-trained FFM models, no need to start from scratch; simply prepare the data to begin training your customized model.
✓ Lower Entry Barriers: No need to build AI supercomputers; start training LLM from thousands of yuan.
✓ Simple and Fast: One-stop service for quickly deploying fine-tuned models in cloud.
✓ Two fine-tuning strategies: Offering full-parameter and LoRA fine-tuning strategy. Choose the suitable one to create your own exclusive model!
*Please refer to AFS Platform & AFS Cloud for complete products and prices.
〝Enterprise Brain Computing Plan〞 Specially designed for those needing massive computing power ! The longer the contract, the greater the savings.
| Plan | Contract duration | Price (NTD, Tax Exclusive) | Annual payment (NTD, Tax Exclusive) | Total usable computing power (NTD, Tax Exclusive) |
|---|---|---|---|---|
| Enterprise Brain 95 | One year | 950,000 | 950,000 | 1,090,000 *Save NT$140,000* |
| Enterprise Brain 95 | Three year | 2,850,000 | 950,000 | 3,420,000 *Save NT$570,000* |
| Enterprise Brain 150 | One year | 1,500,000 | 1,500,000 | 1,800,000 *Save NT$300,000* |
| Enterprise Brain 150 | Three year | 4,500,000 | 1,500,000 | 5,700,000 *Save NT$1,200,000* |
| Enterprise Brain 250 | One year | 2,500,000 | 2,500,000 | 3,300,000 *Save NT$800,000* |
| Enterprise Brain 250 | Three year | 7,500,000 | 2,500,000 | 10,500,000 *Save NT$3,000,000* |
*Please refer to the Pricing for complete products and prices.
企業級大模型優化方案
AFS Shuttle
大型語言模型優化共乘服務
低成本快速投入，概念性驗證 (Proof of Concept，POC) 的最佳選擇
產品功能特色
✓ 容易上手的 No-code Portal
✓ 多種預訓練模型可供選擇
✓ 根據客戶需求和可讀取 token 總量排定製程
✓ 國家級的資安防護與租戶環境隔離
✓ 能進一步與安全線路環境整合(含 VPN )
✓ 提供包套固定價格給企業用戶快速取得 GPU 資源進行 POC 驗證
Free Consultation Service
Contact our experts to learn more and get started with the best solution for you.