Consumption
Generally. The TextQL Virtual Sandcastle Service is a cloud data platform provided by TextQL (“TextQL”, “we”, “us”, “our”) to TextQL customers (each a “Customer”, “you”, “your”) as a service which consumes resources for distinct functions as set forth herein. The service provides Agentic Analytics capabilities through Ana, our AI Analyst, and is available in different service tiers and deployment options. Customer queries are processed using virtual compute resources and AI inference capabilities, charged based on Agent Compute Units (ACUs) and token consumption.Agent Compute Units (ACUs)
Compute (Virtual Sandcastle Service). TextQL bills for compute resources using purchasable Agent Compute Units, as described herein (“ACUs” or “Agent Compute Units”). Virtual compute instances use ACUs at certain rates based on the effective size of the compute instance while the applicable instance is in operation, as set forth in the ACU Tables below. Sandcastle Management. Customers can configure a number of reserved sandcastles that are always allocated to ensure immediate availability. When additional capacity is needed beyond the minimum configuration, additional sandcastles are automatically provisioned as required. These additional sandcastles are deallocated once they are no longer reserved. Each agent thread reserves a sandcastle for one hour after the last message to allow for continued responses. AI Inference. TextQL offers AI inference capabilities powered by large language models. AI inference is charged based on the number of input and output tokens processed, as set forth in the VSS Inference Table below. Token consumption is calculated on a per-token basis for both input and output tokens and is billed in ACUs.Standard
The base service tier providing full compute and AI inference capabilities at $2.00 per 1,000 ACUs.
Enterprise
Enhanced service tier with priority support and advanced features at $3.00 per 1,000 ACUs.
Business Critical
Dedicated cloud deployment with enhanced security and compliance features at $4.00 per 1,000 ACUs.
White Label
Fully customizable solution with white-label branding capabilities at $5.00 per 1,000 ACUs.
ACU Consumption Pricing
Service Tier Pricing
| Service Tier | ACU Rate (per 1,000 ACUs) |
|---|---|
| Standard | $2.00 |
| Enterprise | $3.00 |
| Business Critical | $4.00 |
| White Label | $5.00 |
Compute
| Sandcastle Size | VSS (ACUs Per Hour) |
|---|---|
| XS | 500 |
| S | 1,000 |
| M | 2,000 |
| L | 4,000 |
| XL | 8,000 |
| 2XL | 16,000 |
| 3XL | 32,000 |
Inference
| Model | Token Range | Input | Cache Read | Ephemeral Cache | Persistent Cache | Output |
|---|---|---|---|---|---|---|
| MODEL_HAIKU_3 | N/A | 137.5 | 16.5 | 165 | 275 | 687.5 |
| MODEL_HAIKU_3_5 | N/A | 440 | 44 | 550 | 880 | 2,200 |
| MODEL_SONNET_3_5 | N/A | 1,650 | 165 | 2,062.5 | 3,300 | 8,250 |
| MODEL_SONNET_3_7 | N/A | 1,650 | 165 | 2,062.5 | 3,300 | 8,250 |
| MODEL_SONNET_4 | ≤200K | 1,650 | 165 | 2,062.5 | 3,300 | 8,250 |
| MODEL_SONNET_4 | >200K | 3,300 | 330 | 4,125 | 6,600 | 12,375 |
| MODEL_OPUS_4 | N/A | 8,250 | 825 | 10,312.5 | 16,500 | 41,250 |
| MODEL_GPT_4 | N/A | 16,500 | N/A | N/A | N/A | 33,000 |
| MODEL_GPT_4_TURBO | N/A | 5,500 | N/A | N/A | N/A | 16,500 |
| MODEL_GPT_4O | N/A | 1,375 | 687.5 | N/A | N/A | 5,500 |
| MODEL_GPT_4O_MINI | N/A | 82.5 | 41.25 | N/A | N/A | 330 |
| MODEL_GPT_4_1 | N/A | 1,100 | 275 | N/A | N/A | 4,400 |
| MODEL_GPT_4_1_MINI | N/A | 220 | 55 | N/A | N/A | 880 |
| MODEL_GPT_4_1_NANO | N/A | 55 | 13.75 | N/A | N/A | 220 |
| MODEL_O_1 | N/A | 8,250 | 4,125 | N/A | N/A | 33,000 |
| MODEL_O_1_MINI | N/A | 605 | 302.5 | N/A | N/A | 2,420 |
| MODEL_O_3 | N/A | 1,100 | 275 | N/A | N/A | 4,400 |
| MODEL_O_3_MINI | N/A | 605 | 302.5 | N/A | N/A | 2,420 |
| MODEL_O_4_MINI | N/A | 605 | 151.25 | N/A | N/A | 2,420 |
| MODEL_KIMI_K2_INSTRUCT | N/A | 330 | N/A | N/A | N/A | 1,375 |
| MODEL_QWEN3_CODER | N/A | 247.5 | N/A | N/A | N/A | 990 |
| MODEL_QWEN3_CODER_SMALL | N/A | 82.5 | N/A | N/A | N/A | 330 |
| MODEL_GPT_OSS | N/A | 82.5 | N/A | N/A | N/A | 330 |
| MODEL_GPT_OSS_SMALL | N/A | 38.5 | N/A | N/A | N/A | 165 |
Additional Services
| Service Category | Type | ACUs / Unit |
|---|---|---|
| Storage | Standard | 0 / GB / month |
| Connectivity | Ingress | 0 / GB |
| Connectivity | Egress | 0 / GB |
| Tools | Web Search | 50 / cell |
Deployment Options
Virtual Sandcastle Service (VSS). Standard cloud deployment providing full compute capabilities from XS to 3XL instances with managed infrastructure and AI inference services.Reserved Instances
TextQL offers Reserved Instance pricing for customers who commit to a specific amount of compute capacity for a fixed term. Reserved Instances provide significant discounts on compute ACUs in exchange for an upfront commitment.Reserved Instance Terms and Discounts
Reserved Instances are available for 1-year, 2-year, and 3-year terms with the following discount rates applied to compute ACUs only:| Term Length | Discount | Effective Compute ACU Rate (per 1,000 ACUs) | ||
|---|---|---|---|---|
| Standard | Enterprise | |||
| On-Demand | 0% | $2.00 | $3.00 | |
| 1-Year Reserved | 72% | $0.56 | $0.84 | |
| 2-Year Reserved | 75% | $0.50 | $0.75 | |
| 3-Year Reserved | 78% | $0.44 | $0.66 |
Reserved Instance Details
Scope of Discount. Reserved Instance discounts apply exclusively to:- Virtual Sandcastle compute resources (all sizes from XS to 3XL)
- Minimum sandcastle reservations
- AI Inference tokens (Premium and Standard tiers)
- Storage
- Data transfer (Ingress/Egress)
- Tool usage (Web Search and other tools)
- Non-cancellable during the term
- Non-transferable between accounts
- Non-refundable
- Expandable through additional Reserved Instance purchases
This TextQL Virtual Sandcastle Service Consumption Table may be updated from time to time. Changes shall be effective on the date that TextQL announces they are effective. Any capitalized terms used but not defined herein shall have the meaning set forth in the Agreement or the Documentation, as applicable.