Skip to main content

TWAI Logo

Price List

☑ Multiple payment methods ☑ No up-front cost ☑ No termination fees

Version: 20250425v1.0

Distributor recruitment is underway!
If you are interested in becoming a distributor for TWAI, please contact our dedicated service at sales@twcloud.ai.

  Buy Now  

Per-second Billing

With per-second billing, you don't have to worry about being charged for the unused time in an hour, and you can maximize your investment in testing and development.⁕

⁕ Calculation example:

The time usage of a c.super instance is 11 hours 35 minutes 10 seconds = 41710 seconds = 11.5861 hours. The list credits for on-demand plan is 60.90 Credits/hour. Therefore, the cost is 11.5861 * 60.90 = 705.5934, a total of 705.5934 credits.

Notes

  • Billing Unit: TWAI Wallet Credits
  • Actual charges are subject to TWAI Member Center billing.
  • TWAI reserves the right to change prices at any time. All price changes will be effective after announcements to customers.

    • U-T Cloud​

    TWAI Cloud​

    AI Foundry Service (AFS)​

    AFS ModelSpace​

    Public Mode​

    • AFS ModelSpace enables you to deploy native large language models and Formosa Foundation Model series.
    • If you use the AFS ModelSpace with an AFS project account, the charges will be billed under the AFS project account.
    • See more
    Model series⁕Type
    (Input / Output)    		Credits⁕⁕
    (Per thousand tokens)
    Llama3.3-FFMAFS.MS.inp.Llama3.3-FFM-70B-32K0.0315
    AFS.MS.otp.Llama3.3-FFM-70B-32K
    Llama3.2-FFMAFS.MS.inp.Llama3.2-FFM-11B-V-32K⁕⁕⁕⁕0.0105
    AFS.MS.otp.Llama3.2-FFM-11B-V-32K
    Llama3.1-FFMAFS.MS.inp.Llama3.1-FFM-70B-32K0.0315
    AFS.MS.otp.Llama3.1-FFM-70B-32K
    AFS.MS.inp.Llama3.1-FFM-8B-32K0.00525
    AFS.MS.otp.Llama3.1-FFM-8B-32K
    FFM-embedding-v2.1AFS.MS.inp.FFM-embedding-v2.1⁕⁕⁕0.0021
    FFM-embedding-v2AFS.MS.inp.FFM-embedding-v2⁕⁕⁕0.0021
    TAIDEAFS.MS.inp.TAIDE-LX-7B0.00525
    AFS.MS.otp.TAIDE-LX-7B
    Qwen2.5-coderAFS.MS.inp.Qwen25-Coder-7B-32K0.00525
    AFS.MS.otp.Qwen25-Coder-7B-32K
    Jina-embeddings-v2AFS.MS.inp.JINA-embeddings-v2-base-code⁕⁕⁕0.0021
    BAAI-bge-rerankerAFS.MS.inp.bge-reranker-v2-m3⁕⁕⁕0.0021
    info

    ⁕Model names with "FFM" indicate the traditional Chinese-enhanced new weights by TWAI.
    ⁕⁕Public mode is billed by tokens, per thousand tokens as a billing unit, and it will be calculated as 1,000 tokens when the usage is less than 1,000 tokens.
    ⁕⁕⁕Both the input and output token usage of the model will be incorporated into the input product measurement billing.
    ⁕⁕⁕⁕The input token of a visual-language model includes the tokens converted from image files. For the conversion formula and rules, please refer to the Token Calculation for Visual-Language Models.


    T-Proof Service​

    • Provides data notarization services that leverage the tamper-proof and traceable features of blockchain to enhance the legal validity of data.
    • Data is uploaded to the blockchain every ten minutes, and will be charged on a per-upload basis.
    • See more
    ServicetypeSpecCredits
    (Per upload)
    Notarization servicetpf.notarize.stdStandard5.25
    Verification servicetpf.verify.stdStandard0.00⁕
    Retrieval servicetpf.retrieve.stdStandard0.00⁕
    info

    ⁕ Verification and Retrieval services are offered a discount price until December 31, 2024.


    Compute​

    Virtual Compute Service (VCS)​

    • Supports most common Linux distributions (including Ubuntu, CentOS, Rocky) and Windows.
    • The CPU instance series is equipped with Intel® Xeon® data center level processors.
    • The GPU instance series is equipped with NVIDIA® Tesla® V100 and A100 GPUs, accelerating AI training, inference, and high-performance computing.
    • VCS offers several sizes of system disks for your instances. Moreover, you can attach data disks to instances to grow your storage space flexibly with Virtual Disk Service. Also, we provide image service for the use of instance replication. See Virtual Disk Service (VDS) for more pricing information.
    • Provide static IPs and floating IPs. See Elastic IP (EIP) for more pricing information.
    • See more
    CPU instance seriesInstance typevCPUMemory
    (GB)
    Credits
    (Per hour)
    Linux⁕

    		Windows Server ♣
    		 SQL Server Web♥
    v-seriesv.super2163.154.9357.224
    v.xsuper4326.38.242514.4375
    v.2xsuper86412.6016.48528.875
    v.4xsuper1612825.2032.9757.75
    v.8xsuper3225650.4065.94115.50
    v1-seriesv1.super221.9953.20255.313
    v1.xsuper443.996.40510.626
    v1.2xsuper887.9812.8121.252
    v1.4xsuper161615.9625.6242.504
    v1.8xsuper323231.9251.2485.008
    v2-seriesv2.super242.15253.3815.544
    v2.xsuper484.3056.646511.088
    v2.2xsuper8168.6113.282522.176
    v2.4xsuper163217.2226.56544.352
    v2.8xsuper326434.4453.1388.704
    v4-seriesv4.super282.523.6965.9535
    v4.xsuper4165.047.402511.8965
    v4.2xsuper83210.0814.80523.793
    v4.4xsuper166420.1629.6147.586
    v4.8xsuper3212840.3259.2295.172
    GPU instance seriesInstance typeGPU (pcs)GPU typevCPUMemory
    (GB)
    Credits
    (Per hour)
    Linux⁕

    		Windows Server♣
    vgv-seriesvgv.xsuper1NVIDIA Tesla V100/32GB89073.5076.65
    vgv.2xsuper216180147.00153.30
    vgv.4xsuper432360294.00306.60
    vgv.8xsuper864720588.00613.20
    vgah-series
    (Special Supply)♠    		vgah.xsuper1NVIDIA A100/40GB SXM30120101.85N/A
    vgah.2xsuper260240203.70N/A
    vgah.4xsuper4120480407.40N/A
    vgah.8xsuper8240960814.80N/A
    vga1-series♦vga1.8xsuper8NVIDIA A100/40GB SXM2409601060.50N/A
    info

    ⁕GPU 加速運算個體：提供專門處理圖形任務的 GPU 核心，能快速完成大量運算。
    如何取得 GPU 資源？ 請參考 FAQ-資源監控與配置-Q4 之說明。

    INFO

    ⁕ Ubuntu, CentOS, and Rocky Linux are provided.
    ♣ Include the license for Windows Server Standard edition.
    ♥ Include the licenses for Windows Server Standard edition and SQL Server Web edition.
    ♦ The vga1-series equips with local 11TB SSD and is not available on demand. If you have any needs, please contact sales@twcloud.ai.
    ♠ The vgah series are special supply products. Inquiries about this series, please contact sales@twcloud.ai to apply.


    Virtual K8s Service (VKS)​​

    • Provides Kubernetes cluster services.
    • Compute node specifications support various VCS instance types, including the vgv series.
    • See more
    Compute resourcesTypeCredits
    (Per hour)
    Cluster serviceVKS.base.std2.73
    v-seriesVKS.v seriesSame as VCS v-series
    v1-seriesVKS.v1 seriesSame as VCS v1-series
    v2-seriesVKS.v2 seriesSame as VCS v2-series
    v4-seriesVKS.v4 seriesSame as VCS v4-series
    vgv-seriesVKS.vgv seriesSame as VCS vgv-series
    Network & securityTypeCredits
    Network Flow (NFL)NetFlow_CHF_out3.15 /GB
    Elastic IP (EIP)sip.renstatic105.00 /IP-month
    sip.entstatic105.00 /IP-month
    Load balancingTypeCredits
    (Per hour)
    Virtual K8s Service(VKS)VKS.lbs.basic0.63
    StorageTypeCredits
    Virtual K8s Service(VKS)VKS.hdd.std1.26 /GB-month
    Virtual K8s Service(VKS)VKS.ssd.std2.73 /GB-month
    info

    Compute nodes are provided with the Ubuntu OS.


    Cloud PC (CPC)​

    • Equipped with Intel® Xeon® Gold 61 series processors, the maximum memory processing speed is 2666 MHz.
    • Built-in 200 GB HDD system disk.
    • The OS is Windows Server 2019, with license.
    • Pre-installed ASUS WebStorage - A.I. Secure Drive software. You can log in to the software with your WebStroage member account and use it directly.
    • See more
    Instance seriesInstance typevCPUMemory
    (GB)    		Credits
    (Per hour)
    cpc-seriescpc.c2m4242.625

    Storage​

    Cloud Object Storage (COS)​

    • COS provides highly secure and durable storage; you can store and access various data types over the Internet. COS can be used as a general data storage space independently and is suitable for data storage and transmission on the TWAI platform.
    • The bandwidth cost of data transfer is included in the storage pricing and won't incur an additional charge.
    • See more
    SpecCredits
    (GB-month)
    Standard object storage0.63

    Virtual Disk Service (VDS)​

    • VDS is designed for TWAI VCS and provides highly available, consistent, low-latency storage service.
    • Services: VCS instance ⁕system disks, data disks, and instance backup images.
    • VDS provides two strorage options: Solid State Drive (SSD) and Hard Disk Drives (HDD)
      • Solid State Drives (SSD)
        • Provides low-latency service (less than 10 ms) and consistent I/O performance with significantly high throughput (from 3 IOPS/GB to 16 IOPS/GB). SSD strengthens reliability and scalability for applications, and it is best suited for workloads with high throughput and enterprise applications that require high-performance.
        • Provides a maximum throughput of 1750 MB/s for each VCS instance.
        • Supports storage size up to 16 TB.
      • Hard Disk Drives (HDD)
        • Provides a maximum throughput of 1750 MB/s for each VCS instance. HDD is best suited for large datases that are infrequently accessed and are not latency-sensitive.
        • Provides a maximum throughput of 1750 MB/s for each VCS instance.
        • Supports up to 500 GB of storage space as a system disk.
        • Supports up to 16 TB of storage space as a data disk.
    • See more
    Disk typeSpecCredits
    (GB-month)
    Solid State Drives (SSD)Standard2.73
    Hard Disk Drives (HDD)Standard1.26
    Image static storage⁕⁕Basic0.63
    info

    ⁕ We offer free storage of the first 100 GB for system disks.
    ⁕⁕ For VCS image static storage. For VCS image in the workspace, please refer to Hard Disk Drives (HDD).

    Networking & Security​

    Load Balancing Service (LBS)​

    • Load Balancing Service supports application-layer (HTTP and HTTPS) and network-layer (TCP) data transfer protocols, allowing you to choose the appropriate load balancing method to maintain service flexibility, scalability and high availability.
    • See more
    SpecCredits
    (Per hour)
    Classic0.63
    (Billing start date: 7/7/2022)

    Elastic IP (EIP)​

    • Static IPs are project-wise resources. You can freely assign IPs to different resources in your project.                
    • Applicable services: VCS and LBS.
    • Default subscription limit per project: 5 static IPs.
    • The billing starts once the subscription is successful. The day you unsubscribe will be billed.*If you subscribe and unsubscribe the same day, you will be billed for the actual usage period.**
    • See more
    TypeModelSpecUnitPrice
    Subscriptionsip.entstaticUsage fee of Static IP - Commercial Networkper IP-month105.00
    sip.renstaticMaintainence fee of Static IP - Research and Education Network105.00
    INFO

    ⁕ Example: Subscribe at 11/26 15:00, and cancel subscription at 11/30 21:00. Billing period: 11/26 15:00:00 - 11/26 21:00:00.

    ⁕⁕ Example: Subscribe at 11/26 15:00, and cancel subscription at 11/30 21:00. Billing period: 11/26 15:00:00 - 11/26 21:00:00.


    Data Transfer: Virtual Compute Service (NetFlow)​

    • There is no charge for data ingress traffic, and data transferring between VCS, only charges for data egress traffic!
    • The monthly billing is based on the total outbound data transfers by all VCS instances.                            
    Transfer directionCredits
    (GB)
    Data transfers in to TWAI VCS from InternetFree
    Data transfers out from TWAI VCS to Internet3.15
    Data transfers between TWAI VCSFree

    Advanced Virtual Firewall (VFW)​

    • Advanced Virtual Firewall provides commercial firewall services. You can deploy firewall in the cloud and manage the security policies.
    • See more
    ModelSpecCredits
    (Per hour)
    pa_byolStandard8.61