ArthenServePrivate Flagship
Published Before Enrollment
- Input
- Finalizing
- Output
- Finalizing
Compare the Whole Relationship
Price Is One Column.
Trust Is the Rest.
Commercial API terms vary by product and contract.
This table compares published standard terms.
| What Matters | ArthenServePrivate Service | OpenAIAPI | AnthropicAPI / Commercial | DeepSeekPublished Services |
|---|---|---|---|---|
| Privacy Architecture | Privacy-First by DefaultContent privacy IS the service design. Your content is for your eyes only. | Enterprise controls available, plenty of asterisks. | Commercial controls available, do you trust them? | Standard hosted-service model, but under the watchful eye of the Chinese government... |
| Default AI Content Retention | No Content ArchiveYour prompts, files, and results are unmonitored and not retained after processing. | Abuse-monitoring logs may retain content up to 30 days. Some features retain application state. | API inputs and outputs are normally deleted within 30 days. Policy-flagged content may be held longer. | Inputs and chat history may be retained while the account exists. |
| Training on Your Input/Output Data by Default | Nope! NeverNever ever. You can't even opt-in by accident. | Yes, it's opt-out by default unless you pay for various enterprise/API options. | Yes, it's opt-out for consumer user data. It's opt-in for commercial customers. | Yes. Their policy permits use to train. Opt-out rights may apply, but who knows if they really comply? |
| Uncensored Model Options | AbsolutelyNo outside model vendor’s moderation layer. | No. Provider policies and safety systems apply. | No. Provider policies and safety systems apply. | No. Provider policies and content controls apply. |
| Inference Hardware | Owned and operated by the Arthenfold. | Provider and partner infrastructure options | First-party API plus AWS, Google Cloud, and Azure options | Provider systems and disclosed service providers |
| Processing Location | Arthenfold HQ, USA.No routing to an outside AI provider. | Global processing with eligible regional controls at added cost. | Global by default. US inference geography and cloud-region options at added cost. | Policy states personal data is directly processed and stored in China. |
| Operational Record Scope | Content-free service recordsUsage metrics (not content), timing, model choice, status, and account identifiers. No input/output content is recorded. | Metadata, classifier outputs, and monitored customer input/output content. | Usage, safety, commercial service records, and monitored customer input/output content. | Device, network, usage, diagnostic, account, and content data. |
| Who Supports You | US-based human support with crystal clear American English accents. | Scaled provider support. | Scaled provider support. | Scaled provider support. |
“Uncensored” does not mean lawless. ArthenServe is intended for lawful work and may use content-free account, rate, and security controls. It means eligible requests are not screened to enforce an outside model vendor’s editorial policy. Specialized clients or applications can still apply their own safeguards.
A Public Price Reference
The Token Market.
In Context.
Selected flagship public API rates per one million short-context tokens, checked September 2, 2026. These models are not equivalent in quality, speed, tokenizer, or capabilities.
OpenAIGPT-5.6 Sol
Standard API
- Input
- $4.00 / 1M Tokens
- Output
- $20.00 / 1M Tokens
AnthropicClaude Fable 5.1
Standard API
- Input
- $10.00 / 1M Tokens
- Output
- $50.00 / 1M Tokens
DeepSeekDeepSeek V4 Pro
Peak API Rate
- Input
- $1.32 / 1M Tokens
- Output
- $3.96 / 1M Tokens
We may not be the cheapest token.
We are the only PRIVATE token.The value is knowing whose hardware processed your work, knowing your content was not retained, and knowing it will not become someone else’s training material.Two Ways to Buy
Stay Flexible.
Or Reserve Capacity.
Start light, commit when it makes sense,
and choose exactly what happens at your limit.
01 / MeteredLowest Commitment
Pay as You Go
A simple monthly account for variable or occasional work. Your bill follows measured usage—nothing to estimate in advance.
Your Usage+Service Margin=Your Bill
- No monthly token commitment
- Transparent per-service rates
- Usage dashboard and spend controls
- Same privacy architecture as every plan
02 / SubscriptionBest Ongoing Value
Private Capacity
Choose one service or combine them. A defined monthly budget lowers the effective usage rate and makes recurring work easier to plan.
Monthly Budget+Your Overage Rule=No Surprises
- Service-specific or inclusive plans
- Published monthly token or generation budget
- Preferred rates above the included budget
- Continue, confirm, or stop at the limit
Build Your Subscription
One Service.
Or the Whole Workshop.
Each subscription reserves a defined monthly allowance.
We size it with you before enrollment.
Language & Code
Private Reasoning
Monthly Token Budget
- Writing and analysis
- Programming assistance
- Open-weight model selection
Image & Design
Private Generation
Monthly Generation Budget
- Image creation
- Visual exploration
- Private workflow options
Speech & Audio
Private Transcription
Monthly Minute Budget
- Speech to text
- Caption preparation
- Batch processing
ArthenServe Complete
The Whole Workshop
Shared Multi-Service Budget
- Language, image, and speech
- One consolidated allowance
- Priority capacity planning
Draft Pricing Structure Exact dollar rates, included budgets, and model availability will be published after capacity and power-cost benchmarking. No customer will be asked to enroll without a clear rate sheet.
You Control the Limit
Your Budget Ends.
Your Choice Begins.
A token budget should make spending predictable—not create an interruption trap. Choose the behavior that fits the way you work, and change it for a future billing period.
- 01Automatic
Keep Going
Automatically continue at the published pay-as-you-go rate.
- 02Approval
Ask Me First
Pause new work and wait for you to approve additional usage.
- 03No Overage
Hard Stop
Decline new requests at the limit. No overage and no surprise bill.
Verify the Comparison
Read the Published Terms.
Competitor practices change. We cite the primary pages behind this draft and will date every published comparison.
Straight Answers
Know What
You’re Buying.
Need a specific capacity or privacy arrangement?
Ask Us Directly
When Will Exact Prices Be Available?
Before paid enrollment. We’re benchmarking real throughput, power use, concurrency, model mix, and support cost so the rate sheet is sustainable and honest. This page currently describes the pricing architecture, not a binding offer.
How Does Pay as You Go Work?
Your account records content-free usage totals and bills at the published rate for the service and model used. There is no monthly token commitment. Minimums, billing cadence, and payment terms will be disclosed before enrollment.
What Happens When My Subscription Budget Runs Out?
You choose in advance: continue automatically at the published metered rate, pause and ask for approval, or stop new requests until the budget resets. The final customer dashboard should make that setting explicit and easy to change.
Does ArthenServe Keep My Prompts or Results for Billing?
No. Billing and service operations need totals, timestamps, model selection, status, and account identifiers—not the content of your prompts, files, or results. Read the full distinction in the ArthenServe data-handling promise.
What Does “Uncensored” Mean Here?
ArthenServe can offer eligible open-weight models without an outside vendor’s prompt moderation layer for lawful use. It does not waive applicable law, guarantee every model will answer every prompt, or prevent your chosen client from applying its own safeguards.
Are Tokens the Right Unit for Every Service?
No. Language models are naturally metered in input and output tokens. Transcription is clearer by audio minute, while image and video work may use generations, steps, resolution, or compute time. Every rate sheet should use the unit that customers can actually understand.
Can I Get Dedicated Capacity or a Team Plan?
That belongs in the bespoke tier. Tell us about concurrency, model, throughput, retention, and support requirements. We’ll confirm whether dedicated capacity is practical and provide written scope before any commitment.
ArthenServe by Arthenfold
Buy the Compute.
Keep the Confidence.
Tell us what you want to run, how often you expect to use it,
and where uninterrupted service matters most.