OpenAI0.08 ₽ / request
GPT-5.6 (gpt-5.6-luna)
A fast, economical model for everyday questions and short work tasks.
Fast answersRewritingConcise explanations
Context128,000 tokens
Privacy ModeZDR (Zero Retention)
AI Model Catalog
Curated Selection. Honest Specialization.
We only host models with distinct strengths, proven reliability, and clear pricing. No maze of identical variants — every model has a verified purpose, context window, and privacy factsheet.
01 — Catalog Grid
Every Model Suited for Its Purpose
OpenAI0.24 ₽ / request
GPT-5.6 (gpt-5.6-terra)
A balanced default for analysis, coding, and complex instructions.
AnalysisCodingDocument work
Context200,000 tokens
Privacy ModeZDR (Zero Retention)
OpenAI0.36 ₽ / request
GPT-5.6 (gpt-5.6-sol)
The flagship choice for difficult reasoning, research, and high-stakes work.
Deep analysisResearchComplex architecture
Context200,000 tokens
Privacy ModeZDR (Zero Retention)
Anthropic0.42 ₽ / request
Claude 5 (claude-sonnet-5)
A strong all-round model for careful writing, coding, and long context.
WritingCodingLarge documents
Context200,000 tokens
Privacy ModeZDR (Zero Retention)
Anthropic0.65 ₽ / request
Claude 5 (claude-opus-5)
A premium model for complex planning, analysis, and long autonomous tasks.
StrategyAgentsDeep editing
Context200,000 tokens
Privacy ModeZDR (Zero Retention)
Google0.15 ₽ / request
Gemini 3 (gemini-3.6-flash)
A fast multimodal model for large files, images, and high-volume work.
ImagesLong contextFast processing
Context1,000,000 tokens
Privacy ModeZDR (Zero Retention)
Google2.50 ₽ / image
Gemini Image (nano-banana-pro)
Premium image generation and editing with strong prompt adherence.
Image generationEditingText in images
ContextMultimodal tokens
Privacy ModeStandard 90d
xAI0.29 ₽ / request
Grok 4 (grok-4.5)
A general model with strong reasoning and current search tools.
ReasoningCurrent topicsTechnical questions
Context128,000 tokens
Privacy ModeZDR (Zero Retention)
DeepSeek0.05 ₽ / request
DeepSeek V4 (deepseek-v4-flash)
A fast, very cheap model for coding, drafts, and high-volume work.
Fast codingDraftsHigh-volume tasks
Context1,000,000 tokens
Privacy ModeZDR (Zero Retention)
DeepSeek0.08 ₽ / request
DeepSeek V4 (deepseek-v4-pro)
A cost-efficient strong model for coding, mathematics, and technical analysis.
CodingMathematicsTechnical analysis
Context128,000 tokens
Privacy ModeZDR (Zero Retention)
Alibaba0.22 ₽ / request
Qwen 3.8 (qwen-3.8-max)
A long-context multimodal model for complex documents, video, and code.
Long documentsVideoCoding
Context1,000,000 tokens
Privacy ModeZDR (Zero Retention)
Moonshot AI0.18 ₽ / request
Kimi K3 (kimi-k3)
A strong model for coding, agent workflows, and long-form research.
AgentsCodingResearch
Context2,000,000 tokens
Privacy ModeZDR (Zero Retention)
MiniMax0.10 ₽ / request
MiniMax M3 (minimax-m3)
A cost-efficient long-context model for daily work and automation.
AutomationLong contextDaily tasks
Context1,000,000 tokens
Privacy ModeZDR (Zero Retention)
Zhipu AI0.22 ₽ / request
GLM 5 (glm-5.2)
The GLM flagship for coding, agents, and long multi-step work.
CodingAgentsLong-horizon tasks
Context1,000,000 tokens
Privacy ModeZDR (Zero Retention)
— Match to a task
Not sure which model to pick?
Describe the task in plain language — we’ll suggest 1–3 fitting models.
02 — Nontechnical Comparison
How to Choose in 30 Seconds
If you are unsure where to begin, use this concise recommendation summary.
| Your Task | Recommended Model | Key Advantage | Context Limit | Relative Cost |
|---|---|---|---|---|
| Complex Code & Architecture | GPT 5.6 Sol | Highest precision for algorithms and multi-step tasks | 200,000 | Flagship |
| Writing, Editing & Tone | Claude Sonnet 5 | Natural language without robotic cliches and deep nuances | 200,000 | Balanced |
| Huge Documents & Multi-file Analysis | Kimi K3 / Gemini 3.6 | 1M to 2M tokens context window for entire repositories | 1,000,000 - 2,000,000 | Economical |
| Fast Economical Coding | DeepSeek V4 Pro | High speed and extremely low cost per request with high accuracy | 128,000 | Lowest |
| Real-time Web Research & News | Grok 4.5 | Live web search with inline verifiable citations | 128,000 | Balanced |
| Cheap automation & long context | MiniMax M3 / Qwen 3.8 | Long context at a low cost per request | 1,000,000 | Economical |
| Agents & coding with GLM | GLM 5.2 | Strong reasoning with a practical agentic style | 1,000,000 | Balanced |