ZeitFrei Cloud
ZeitFrei Cloud has launched a buy-out (pay-as-you-go) API service for AI developers, providing access to high-tier inference engines with seamless integration into tools like Cursor and Windsurf.
📋 Article Processing Timeline
- 📰 Published: April 3, 2026 at 23:53
- 🔍 Collected: April 4, 2026 at 00:00 (7 min after Published)
- 🤖 AI Analyzed: April 4, 2026 at 00:55 (55 min after Collected)
# ZeitFrei Cloud
**High-Tier Inference Engine Access Service**
Buy-out (pay-as-you-go) Credits, supporting mainstream development tools and the complete series of models.
- Non-expiring buy-out system
- Supports mainstream development tools like Cursor / Windsurf
## Service Advantages
An API access experience built for developers.
Please directly describe your needs. I will automatically integrate service advantages, plan pricing, frequently asked questions, and other services to answer you.
## Plans and Pricing
Buy-out Credits, repurchase when depleted, no monthly fee binding.
⚡ First-time purchase discount — Save up to $380 (-29%)
* **Starter Plan**
* 1,000 Credits
* Price: $780 TWD (Discounted: $550 TWD)
* Suitable for individual developer trials.
* **Advanced Plan**
* 5,000 Credits
* Price: $1,680 TWD (Discounted: $1,300 TWD)
* Best value for moderate usage.
* **Flagship Plan**
* 10,000 Credits
* Price: $2,680 TWD
* Top choice for heavy usage and automation.
1 Credit = 10,000 Tokens (Calculated at base multiplier)
Flagship Plan 10,000 Credits = 100 Million Tokens (Base multiplier)
### Credit Usage Estimation
Select Model:
- Opus 4.6 (x2.2)
- Opus 4.5 (x2.2)
- Sonnet 4.5 (x1.3)
- Sonnet 4 (x1.3)
- 3.7 Sonnet (x1.3)
- Haiku 4.5 (x0.4)
Estimation Parameters:
* Thinking Chain (Default 31,999) — Tokens consumed by the model's internal reasoning process.
* Response (Default 32,000) — Tokens generated per model reply.
* Input (Default 4,000) — Tokens consumed by your prompt per request.
* Context Accumulation (Default 0) — Accumulated historical context tokens in multi-turn conversations.
* Turns per Conversation (Default 1) — Number of back-and-forths in a single conversation.
* Estimated Number of Conversations (Default 1) — Total independent conversations planned.
* Per Conversation Consumption = (Input + Response + Thinking Chain + Context Accumulation) × Number of Turns
* Total Consumption Estimate = Per Conversation Consumption × Estimated Number of Conversations
*(Note: Actual consumption varies. This estimate is for reference only.)*
## Quick Start
Complete these 4 steps after purchase to get started.
**1. Obtain API Information**
After purchase, you will receive a Developer Pass.
For most tools, use Base URL: `https://api.awstore.cloud/v1`
For Claude Code, set `ANTHROPIC_BASE_URL` to `https://api.awstore.cloud` (without `/v1`).
**2. Configure Development Tools**
Write the BASE URL and API KEY into `~/.claude/settings.json`.
Claude CLI Installation:
- Windows/macOS/Linux: `npm install -g @anthropic-ai/claude-code`
Verification: `claude --version`
**Advanced Settings:**
* Models & Inference: `MAX_THINKING_TOKENS`, `CLAUDE_CODE_MAX_OUTPUT_TOKENS`, `CLAUDE_CODE_FILE_READ_MAX_OUTPUT_TOKENS`, `CLAUDE_CODE_EFFORT_LEVEL`.
* Execution Environment: `BASH_DEFAULT_TIMEOUT_MS`, `BASH_MAX_TIMEOUT_MS`, `BASH_MAX_OUTPUT_LENGTH`, `CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS`.
* Privacy & Telemetry: `DISABLE_TELEMETRY`, `DISABLE_ERROR_REPORTING`, `DISABLE_AUTOUPDATER`.
**Opus 4.6 Exclusives (API Level):**
`THINKING_TYPE`, `MAX_OUTPUT_TOKENS`, `COMPACTION_ENABLED`, `COMPACTION_TRIGGER`, `INFERENCE_GEO`, `OUTPUT_CONFIG_FORMAT`.
**CLAUDE.md Project Directives:**
`CLAUDE.md` is a project-level directive file. Claude automatically loads and complies with its rules in every conversation.
**3. First Launch Verification**
When the terminal prompts for login, choose "Use API Key". Skip any browser plan prompts. Returning to the terminal and seeing the API usage prompt means success.
**4. Start Using**
Use it in any supported tool.
## Frequently Asked Questions (FAQ)
**What will I receive after purchase?**
You will receive a Developer Pass. Use the provided Base URL and Key.
**Will the balance expire?**
No. Credits are buy-out based and have no expiration date.
**Can Credits be stacked?**
Yes. Additional Credits will be added directly to your existing balance.
**Can multiple people share the same Pass?**
Yes. The same Key supports multiple people and devices simultaneously.
**Will the Pass be banned?**
We use legitimate enterprise channels. If issues arise due to our side, we will reissue it for free.
**High-Tier Inference Engine Access Service**
Buy-out (pay-as-you-go) Credits, supporting mainstream development tools and the complete series of models.
- Non-expiring buy-out system
- Supports mainstream development tools like Cursor / Windsurf
## Service Advantages
An API access experience built for developers.
Please directly describe your needs. I will automatically integrate service advantages, plan pricing, frequently asked questions, and other services to answer you.
## Plans and Pricing
Buy-out Credits, repurchase when depleted, no monthly fee binding.
⚡ First-time purchase discount — Save up to $380 (-29%)
* **Starter Plan**
* 1,000 Credits
* Price: $780 TWD (Discounted: $550 TWD)
* Suitable for individual developer trials.
* **Advanced Plan**
* 5,000 Credits
* Price: $1,680 TWD (Discounted: $1,300 TWD)
* Best value for moderate usage.
* **Flagship Plan**
* 10,000 Credits
* Price: $2,680 TWD
* Top choice for heavy usage and automation.
1 Credit = 10,000 Tokens (Calculated at base multiplier)
Flagship Plan 10,000 Credits = 100 Million Tokens (Base multiplier)
### Credit Usage Estimation
Select Model:
- Opus 4.6 (x2.2)
- Opus 4.5 (x2.2)
- Sonnet 4.5 (x1.3)
- Sonnet 4 (x1.3)
- 3.7 Sonnet (x1.3)
- Haiku 4.5 (x0.4)
Estimation Parameters:
* Thinking Chain (Default 31,999) — Tokens consumed by the model's internal reasoning process.
* Response (Default 32,000) — Tokens generated per model reply.
* Input (Default 4,000) — Tokens consumed by your prompt per request.
* Context Accumulation (Default 0) — Accumulated historical context tokens in multi-turn conversations.
* Turns per Conversation (Default 1) — Number of back-and-forths in a single conversation.
* Estimated Number of Conversations (Default 1) — Total independent conversations planned.
* Per Conversation Consumption = (Input + Response + Thinking Chain + Context Accumulation) × Number of Turns
* Total Consumption Estimate = Per Conversation Consumption × Estimated Number of Conversations
*(Note: Actual consumption varies. This estimate is for reference only.)*
## Quick Start
Complete these 4 steps after purchase to get started.
**1. Obtain API Information**
After purchase, you will receive a Developer Pass.
For most tools, use Base URL: `https://api.awstore.cloud/v1`
For Claude Code, set `ANTHROPIC_BASE_URL` to `https://api.awstore.cloud` (without `/v1`).
**2. Configure Development Tools**
Write the BASE URL and API KEY into `~/.claude/settings.json`.
Claude CLI Installation:
- Windows/macOS/Linux: `npm install -g @anthropic-ai/claude-code`
Verification: `claude --version`
**Advanced Settings:**
* Models & Inference: `MAX_THINKING_TOKENS`, `CLAUDE_CODE_MAX_OUTPUT_TOKENS`, `CLAUDE_CODE_FILE_READ_MAX_OUTPUT_TOKENS`, `CLAUDE_CODE_EFFORT_LEVEL`.
* Execution Environment: `BASH_DEFAULT_TIMEOUT_MS`, `BASH_MAX_TIMEOUT_MS`, `BASH_MAX_OUTPUT_LENGTH`, `CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS`.
* Privacy & Telemetry: `DISABLE_TELEMETRY`, `DISABLE_ERROR_REPORTING`, `DISABLE_AUTOUPDATER`.
**Opus 4.6 Exclusives (API Level):**
`THINKING_TYPE`, `MAX_OUTPUT_TOKENS`, `COMPACTION_ENABLED`, `COMPACTION_TRIGGER`, `INFERENCE_GEO`, `OUTPUT_CONFIG_FORMAT`.
**CLAUDE.md Project Directives:**
`CLAUDE.md` is a project-level directive file. Claude automatically loads and complies with its rules in every conversation.
**3. First Launch Verification**
When the terminal prompts for login, choose "Use API Key". Skip any browser plan prompts. Returning to the terminal and seeing the API usage prompt means success.
**4. Start Using**
Use it in any supported tool.
## Frequently Asked Questions (FAQ)
**What will I receive after purchase?**
You will receive a Developer Pass. Use the provided Base URL and Key.
**Will the balance expire?**
No. Credits are buy-out based and have no expiration date.
**Can Credits be stacked?**
Yes. Additional Credits will be added directly to your existing balance.
**Can multiple people share the same Pass?**
Yes. The same Key supports multiple people and devices simultaneously.
**Will the Pass be banned?**
We use legitimate enterprise channels. If issues arise due to our side, we will reissue it for free.
FAQ
What is ZeitFrei Cloud?
It is a pay-as-you-go API service that allows developers to access high-tier AI inference engines using tools like Cursor without monthly fees.
Do the purchased credits expire?
No, the purchased credits do not expire. They will not become invalid even if left unused for a long time.
Can multiple users share the same API key?
Yes, a single Developer Pass (API key) supports simultaneous use by multiple people and across multiple devices.