# DevTk.AI

> AI model, API, and developer-tool cost intelligence for builders.

DevTk.AI provides privacy-first tools and guides for model pricing, API spend, rate limits, structured outputs, and MCP workflows. Core tools run client-side in the browser.

## About

DevTk.AI is a focused AI cost intelligence site for developers choosing models, estimating token spend, planning throughput, and shipping MCP/structured-output workflows. Available in English and Chinese.

Key facts:
- Focused cost and API planning tools
- 40+ AI model pricing data
- 25 curated MCP servers
- 47 curated AI tools directory
- 25 curated AI datasets
- Bilingual: English + Chinese
- Updated: May 2026

## Tools

- [AI Token Counter](https://devtk.ai/en/tools/token-counter/): Count tokens for GPT-5, Claude, Llama, Gemini, and more. Real-time tokenization with cost estimation.
- [AI Model Pricing Calculator](https://devtk.ai/en/tools/pricing-calculator/): Compare costs across 40+ AI models. Estimate monthly API spend.
- [LLM VRAM Calculator](https://devtk.ai/en/tools/vram-calculator/): Calculate GPU VRAM requirements for running local LLMs.
- [MCP Manifest Generator](https://devtk.ai/en/tools/mcp-manifest-generator/): Visual builder for Model Context Protocol server configurations.
- [MCP Config Validator](https://devtk.ai/en/tools/mcp-validator/): Validate MCP server configs against the specification.
- [OpenAPI to MCP Converter](https://devtk.ai/en/tools/openapi-to-mcp/): Convert OpenAPI specs into MCP tool definitions.
- [JSON Schema Builder](https://devtk.ai/en/tools/json-schema-builder/): Visual schema builder for LLM structured outputs and function calling.
- [Prompt Diff & Version Compare](https://devtk.ai/en/tools/prompt-diff/): Compare prompt versions side-by-side with token counts.

## Directories

- [AI Tools & Repos](https://devtk.ai/en/directory/ai-tools/): 47 curated AI tools, frameworks, and repositories.
- [MCP Servers](https://devtk.ai/en/directory/mcp-servers/): 25 production-ready MCP servers with install configs.
- [AI Datasets](https://devtk.ai/en/directory/datasets/): 25 curated datasets for LLM training and evaluation.

## AI Model Pricing

DevTk.AI tracks pricing for 40+ AI models including GPT-5, Claude Opus 4.6, Gemini 3.1 Pro, DeepSeek V4, Llama 4, and more. Each model page includes input/output token pricing, context window size, and monthly cost estimates.

- [All Models](https://devtk.ai/en/tools/pricing-calculator/): Full pricing comparison table
- [Model Comparisons](https://devtk.ai/en/compare/openai-vs-anthropic/): Head-to-head provider comparisons

## Blog

- [AI API Pricing Comparison 2026](https://devtk.ai/en/blog/ai-api-pricing-comparison-2026/): Side-by-side pricing for all major AI models.
- [AI Coding Agent Cost Comparison 2026](https://devtk.ai/en/blog/ai-coding-agent-cost-comparison-2026/): Codex, Claude Code, Cursor-style agents, DeepSeek V4, GPT-5.5, and Claude Sonnet cost per task.
- [DeepSeek API Pricing Guide 2026](https://devtk.ai/en/blog/deepseek-api-pricing-guide-2026/): DeepSeek V4 Flash cache-hit pricing, V4 Pro discount, and agent cost math.
- [Claude API Pricing Guide 2026](https://devtk.ai/en/blog/claude-api-pricing-guide-2026/): Opus vs Sonnet vs Haiku pricing, prompt caching, batch API discounts.
- [OpenAI API Pricing Guide 2026](https://devtk.ai/en/blog/openai-api-pricing-guide-2026/): Complete OpenAI API pricing — GPT-5 at $1.25/M, GPT-4.1, o3, batch API discounts.
- [Google Gemini API Pricing Guide 2026](https://devtk.ai/en/blog/gemini-api-pricing-guide-2026/): Gemini 2.5 Pro at $1.25/M, Flash at $0.15/M, free tier details, 1M context pricing.
- [What is MCP? Complete Guide](https://devtk.ai/en/blog/what-is-mcp-guide/): Everything developers need to know about Model Context Protocol.
- [Complete Guide to .cursorrules](https://devtk.ai/en/blog/complete-guide-cursorrules/): Master AI coding rules for Cursor IDE.
- [How to Choose an AI Model](https://devtk.ai/en/blog/how-to-choose-ai-model/): Decision framework for selecting the right LLM.
- [LLM Context Windows Explained](https://devtk.ai/en/blog/llm-context-window-explained/): Understanding token limits and context management.
- [Best Free AI Tools 2026](https://devtk.ai/en/blog/best-free-ai-tools-2026/): Curated list of the best free AI tools for developers.
- [Grok API Pricing Guide 2026](https://devtk.ai/en/blog/grok-api-pricing-guide-2026/): xAI Grok 3 at $3/M, Mini at $0.30/M. Free $25 credit, X/Twitter integration.
- [Mistral API Pricing Guide 2026](https://devtk.ai/en/blog/mistral-api-pricing-guide-2026/): Mistral Large 3 at $2/M, Small 3.1 at $0.20/M. GDPR-compliant, open-weight models.
- [How to Cut AI API Costs by 80%](https://devtk.ai/en/blog/how-to-reduce-ai-api-costs/): 8 proven strategies — prompt caching, batch API, model routing, and more.
- [Claude Opus 4.5 vs GPT-5 for Coding](https://devtk.ai/en/blog/claude-vs-gpt5-coding-comparison-2026/): Head-to-head coding comparison with benchmarks, pricing, and real-world tests.
- [Self-Hosting LLMs vs API: Cost Breakdown 2026](https://devtk.ai/en/blog/self-hosting-llm-vs-api-cost-2026/): GPU costs, breakeven analysis, hidden costs. When to self-host vs use API.
- [AI API Rate Limits Comparison 2026](https://devtk.ai/en/blog/ai-api-rate-limits-comparison-2026/): OpenAI vs Claude vs Gemini vs Grok rate limits. TPM, RPM, tier systems compared.
- [AI Structured Outputs Guide 2026](https://devtk.ai/en/blog/ai-structured-output-guide-2026/): JSON Mode, Function Calling, and Schema validation across OpenAI, Claude, Gemini.
- [Build Your First MCP Server Tutorial](https://devtk.ai/en/blog/build-mcp-server-tutorial-2026/): Step-by-step TypeScript MCP server tutorial with Tools, Resources, and Prompts.
- [Gemini 3.1 Pro Pricing Guide 2026](https://devtk.ai/en/blog/gemini-3-1-pro-pricing-guide-2026/): Google's new flagship — $2.00/M, 77.1% ARC-AGI-2, native video, 1M context.
- [GPT-5.5 in Codex Pricing Guide 2026](https://devtk.ai/en/blog/gpt-5-5-codex-pricing-guide-2026/): OpenAI GPT-5.5, GPT-5.2-Codex, and DeepSeek V4 Flash routing cost guide.
- [AI Video API Pricing 2026](https://devtk.ai/en/blog/ai-video-generation-pricing-2026/): Seedance 2.0, Sora 2, Kling 3.0, Veo 3.1 — per-second API pricing, free tiers, developer guide.

## Reading List

Curated articles from Anthropic on AI safety, developer best practices, and the future of AI:
- [Reading List](https://devtk.ai/en/reading-list/): 18 essential articles from Anthropic's leadership and research team.