Content: text (0 symbol)
Available: 9
Uploaded: 25.10.2026

Positive responses: 3
Negative responses: 0

Sold: 51
Refunds: 0

~ $0.5
The exact amount is on the payment page
🤖 What is this?
It provides access to models via an API for your programs and extensions (OpenCode, Cursor, Openclaw, Hermes, Cline, etc.). 🧠 Available models (Price is in credits per 1 million tokens):
hy3 (input: 198 | output: 792 | cache: 49.5)
minimax-m3 (input: 150 | output: 600 | cache: 30)
mimo-v2.5-pro (input: 217.5 | output: 435 | cache: 1.8)
deepseek-v4-flash (input: 220 | output: 660 | cache: 7)
deepseek-v4-flash-0731 (input: 220 | output: 660 | cache: 7)
deepseek-v4-flash-vision-exp (input: 220 | output: 660 | cache: 7)
kimi-k2.6 (input: 237.5 | output: 1,000 | cache: 40)
kimi-k2.7-code (input: 237.5 | output: 1,000 | cache: 47.5)
glm-5.2 (input: 350 | output: 1,100 | cache: 65)
glm-5.3 (input: 350 | output: 1,100 | cache: 65)
kimi-k2.7-code-highspeed (input: 475 | output: 2,000 | cache: 95)
kimi-k3 (input: 750 | output: 3,750 | cache: 75)
qwen3.8-flash (input: 75 | output: 235)
mimo-v2.6-flash (input: 233.33 | output: 466.67)
glm-5.3-flash (input: 187.5 | output: 625)
deepseek-v4.1-flash (input: 300 | output: 1,200)
deepseek-v4.1-flash-1017 (input: 300 | output: 1,200)
hy4 (input: 396 | output: 1,584)
grok-4.5 (input: 666.67 | output: 2,000)
grok-4.6 (input: 666.67 | output: 2,000)
grok-4.7 (input: 666.67 | output: 2,000)
qwen3.8-max (input: 666.67 | output: 2,000)
deepseek-v4-pro (input: 660 | output: 1,980)
deepseek-v4-pro-0813 (input: 660 | output: 1,980)
minimax-m3.1-flash-preview (input: 1,000 | output: 4,000)
gemini-3.8-flash (input: 1,200 | output: 6,000)
kimi-k3.1-pro (input: 1,333.33 | output: 6,666.67)
gemini-4-argon (input: 1,600 | output: 8,000)

⚡ Please note: cache tokens cost approximately 10 times less than standard input tokens. When reusing context, cache can account for up to 80 of your requests, allowing for significant savings on credits.

💳 Credits
Credits are the service´s internal currency; your balance and the pricing for all models are denominated in them. They are deducted based on the actual volume of text (request + response). There is no expiration date—your key remains active as long as you have credits.

🔌 How does the API work?
This provides unofficial API access to the models. Requests are routed through active, subscribed accounts, making the models available to you via an API.
From your perspective, the experience is virtually identical to using a standard API: simply connect your API key and base URL to OpenCode, Cursor, Openclaw, Hermes, Cline, or your own custom application to interact with the models directly.

📦 What you receive immediately upon purchase:
* 🔑 A personal API key and a Base URL for connection (provided in the instructions).
* 📈 A dashboard to track your token usage.
* 📖 Step-by-step setup guides for various IDEs.
* ⚡ Instant .

💻 Where can it be used?
* In clients such as OpenCode, Cursor, Openclaw, Hermes, Cline, and others (feel free to ask me about other compatible software before purchasing).
* In your own custom bots, scripts, and applications.

⏱ How is the balance deducted?
The tokens you purchase are consumed by the text of your prompts and the AI´s responses. Your key remains active until the entire balance is exhausted—there are no time limits.

⚠️ Important details:
You are purchasing an API key for use with software, not a personal user account. Request limit: ~15 per minute (more than enough for standard coding and work tasks; the limit only becomes noticeable when there is simultaneous load from multiple threads).

🛡 Warranty and Support:
If you run into any issues with the setup process, just reach out—I’ll be happy to help you sort it out.
24.10.2026 10:02:57
взял второй раз, для черновой работы отлично. из минусов - ночью отвалился сервак на пару часов, сейчас продолжает работать дальше. если есть возможность, добавьте minimax m3.1 flash
23.10.2026 8:45:15
работает, на выбор несколько сетей, показывает доступные, нагрузку и прочее. и рабоатет шустро
27.08.2026 3:06:13
всё работает, четко