# ToolKiti - Complete API Reference for AI Agents
> Generated: 2026-07-29 | Total: 32 APIs | Updated daily
> This file is optimized for LLM ingestion. Every API entry includes code examples, pricing, endpoints, and performance metrics.

---

## LLM Providers (大语言模型) — 9 APIs

### [HOT] OpenAI API
- URL: https://toolkiti.org/api/openai
- Description: Access GPT-4o, GPT-4, GPT-3.5, embeddings, DALL-E, Whisper, and TTS models via REST API.
- Pricing: pay-as-you-go (per token)
- Authentication: API Key (Bearer Token)
- Rate Limit: 10000 req/min (varies by tier)
- Latency: <500ms (GPT-4o)
- Max Context: 128K (GPT-4o)
- Popularity: 100/100
- SDKs: Python, Node.js, Go, Java, curl
- Endpoints:
  - https://api.openai.com/v1/chat/completions
  - https://api.openai.com/v1/embeddings
  - https://api.openai.com/v1/images/generations
  - https://api.openai.com/v1/audio/transcriptions
- Key Features:
  - GPT-4o, GPT-4, GPT-3.5 model family
  - Multimodal: vision, audio, text generation
  - Function calling & structured outputs
  - Embeddings API for semantic search
  - Fine-tuning for custom models
- Code Example (bash):
```bash
curl https://api.openai.com/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $OPENAI_API_KEY" \
  -d '{
    "model": "gpt-4o",
    "messages": [{"role": "user", "content": "Hello!"}]
  }'
```

### [POPULAR] Anthropic Claude API
- URL: https://toolkiti.org/api/anthropic
- Description: Access Claude 3.5 Sonnet, Haiku, and Opus models for chat, analysis, and code generation.
- Pricing: pay-as-you-go (per token)
- Authentication: API Key (x-api-key header)
- Rate Limit: 4000 req/min (Tier 4)
- Latency: <800ms (Claude 3.5 Sonnet)
- Max Context: 200K
- Popularity: 87/100
- SDKs: Python, TypeScript, Go, Java, curl
- Endpoints:
  - https://api.anthropic.com/v1/messages
  - https://api.anthropic.com/v1/complete
- Key Features:
  - Claude 3.5 Sonnet, Haiku, Opus models
  - Extended thinking for complex reasoning
  - Computer use (beta) for GUI interaction
  - Vision & document analysis
  - Constitutional AI safety
- Code Example (bash):
```bash
curl https://api.anthropic.com/v1/messages \
  -H "Content-Type: application/json" \
  -H "x-api-key: $ANTHROPIC_API_KEY" \
  -H "anthropic-version: 2023-06-01" \
  -d '{
    "model": "claude-3-5-sonnet-20241022",
    "max_tokens": 1024,
    "messages": [{"role": "user", "content": "Hello!"}]
  }'
```

### [POPULAR] Google Gemini API
- URL: https://toolkiti.org/api/google-gemini
- Description: Access Gemini 2.0 Flash, Pro, and Ultra models with native multimodal and tool-use support.
- Pricing: free tier available, pay-as-you-go
- Authentication: API Key
- Rate Limit: 1500 req/min (paid tier)
- Latency: <600ms (Gemini 2.0 Flash)
- Max Context: 1M (Gemini 1.5 Pro)
- Popularity: 85/100
- SDKs: Python, Node.js, Go, Swift, curl
- Endpoints:
  - https://generativelanguage.googleapis.com/v1beta/models/gemini-2.0-flash:generateContent
- Key Features:
  - Gemini 2.0 Flash, Pro models
  - 1M+ token context window
  - Native multimodal: text, image, audio, video
  - Google Search grounding integration
  - Free tier available
- Code Example (bash):
```bash
curl "https://generativelanguage.googleapis.com/v1beta/models/gemini-2.0-flash:generateContent?key=$GEMINI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "contents": [{"parts": [{"text": "Hello!"}]}]
  }'
```

### [POPULAR] DeepSeek API
- URL: https://toolkiti.org/api/deepseek
- Description: Access DeepSeek-V3 and DeepSeek-R1 reasoning models with competitive pricing.
- Pricing: pay-as-you-go (per token, very low cost)
- Authentication: API Key
- Rate Limit: 500 req/min
- Latency: <1s
- Max Context: 128K
- Popularity: 83/100
- SDKs: Python, curl
- Endpoints:
  - https://api.deepseek.com/v1/chat/completions
- Key Features:
  - DeepSeek-V3 general model
  - DeepSeek-R1 reasoning model
  - Extremely competitive pricing
  - OpenAI-compatible API format
  - 128K context window
- Code Example (bash):
```bash
curl https://api.deepseek.com/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $DEEPSEEK_API_KEY" \
  -d '{
    "model": "deepseek-chat",
    "messages": [{"role": "user", "content": "Hello!"}]
  }'
```

###  Mistral AI API
- URL: https://toolkiti.org/api/mistral
- Description: Access Mistral Large, Small, and Codestral models with efficient architecture.
- Pricing: pay-as-you-go (per token)
- Authentication: API Key
- Rate Limit: 500 req/min
- Latency: <500ms
- Max Context: 128K
- Popularity: 68/100
- SDKs: Python, TypeScript, Go, curl
- Endpoints:
  - https://api.mistral.ai/v1/chat/completions
  - https://api.mistral.ai/v1/embeddings
- Key Features:
  - Mistral Large, Small, Codestral models
  - Efficient architecture for fast inference
  - Open-source model weights available
  - Function calling support
  - Multilingual performance
- Code Example (bash):
```bash
curl https://api.mistral.ai/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $MISTRAL_API_KEY" \
  -d '{
    "model": "mistral-large-latest",
    "messages": [{"role": "user", "content": "Hello!"}]
  }'
```

###  Cohere API
- URL: https://toolkiti.org/api/cohere
- Description: Access Command-R, Embed, and Rerank models for enterprise NLP and search.
- Pricing: free tier available, pay-as-you-go
- Authentication: API Key (Bearer Token)
- Popularity: 65/100
- SDKs: Python, TypeScript, Go, Java, curl
- Endpoints:
  - https://api.cohere.com/v1/chat
  - https://api.cohere.com/v1/embed
  - https://api.cohere.com/v1/rerank
- Key Features:
  - Command-R for business RAG
  - Embed v3 for multilingual embeddings
  - Rerank for search relevance
  - Tool use & multi-step reasoning
  - Enterprise security & compliance
- Code Example (bash):
```bash
curl https://api.cohere.com/v1/chat \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $COHERE_API_KEY" \
  -d '{
    "model": "command-r-plus",
    "message": "Hello!"
  }'
```

###  Groq API
- URL: https://toolkiti.org/api/groq
- Description: Ultra-fast inference API for open-source LLMs using LPU inference engine.
- Pricing: free tier available, pay-as-you-go
- Authentication: API Key
- Rate Limit: 30 req/min (free tier)
- Latency: <200ms (ultra-fast)
- Max Context: 128K
- Popularity: 72/100
- SDKs: Python, curl, OpenAI-compatible SDKs
- Endpoints:
  - https://api.groq.com/openai/v1/chat/completions
- Key Features:
  - LPU inference engine for blazing speed
  - Open-source models: Llama, Mixtral, Gemma
  - Free tier for experimentation
  - OpenAI-compatible API
  - Ideal for real-time applications
- Code Example (bash):
```bash
curl https://api.groq.com/openai/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $GROQ_API_KEY" \
  -d '{
    "model": "llama3-70b-8192",
    "messages": [{"role": "user", "content": "Hello!"}]
  }'
```

###  Together AI API
- URL: https://toolkiti.org/api/together-ai
- Description: Cloud platform for open-source LLM inference with 100+ models including Llama, Mixtral, and DeepSeek.
- Pricing: pay-as-you-go (per token)
- Authentication: API Key
- Popularity: 70/100
- SDKs: Python, TypeScript, curl
- Endpoints:
  - https://api.together.xyz/v1/chat/completions
- Key Features:
  - 100+ open-source models
  - Llama, Mixtral, DeepSeek, Qwen support
  - Inference & fine-tuning endpoints
  - GPU clusters for training
  - OpenAI-compatible API
- Code Example (bash):
```bash
curl https://api.together.xyz/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $TOGETHER_API_KEY" \
  -d '{
    "model": "meta-llama/Llama-3.3-70B-Instruct",
    "messages": [{"role": "user", "content": "Hello!"}]
  }'
```

###  Perplexity API
- URL: https://toolkiti.org/api/perplexity
- Description: Access Sonar search-grounded LLMs with real-time web search and citations.
- Pricing: pay-as-you-go, subscription available
- Authentication: API Key (Bearer Token)
- Rate Limit: 200 req/min
- Latency: <2s
- Max Context: 128K
- Popularity: 58/100
- SDKs: Python, curl, OpenAI-compatible SDKs
- Endpoints:
  - https://api.perplexity.ai/chat/completions
- Key Features:
  - Sonar search-grounded LLMs
  - Real-time web search with citations
  - Up-to-date information retrieval
  - OpenAI-compatible API format
  - Subscription plans for higher limits
- Code Example (bash):
```bash
curl https://api.perplexity.ai/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $PERPLEXITY_API_KEY" \
  -d '{
    "model": "sonar-pro",
    "messages": [{"role": "user", "content": "Latest AI news?"}]
  }'
```

---

## Web Search (搜索服务) — 6 APIs

### [HOT] Tavily Search API
- URL: https://toolkiti.org/api/tavily
- Description: AI-native search engine optimized for LLMs and AI agents. Real-time web data with structured results.
- Pricing: free tier (1000 credits/mo), pay-as-you-go
- Authentication: API Key
- Rate Limit: 100 req/day (free tier)
- Latency: <1s
- Popularity: 95/100
- SDKs: Python, curl
- Endpoints:
  - https://api.tavily.com/search
- Key Features:
  - AI-native search for agents
  - Structured results with summaries
  - Real-time web data extraction
  - Topic-specific deep searches
  - Free tier for development
- Code Example (python):
```python
from tavily import TavilyClient
client = TavilyClient(api_key="$TAVILY_API_KEY")
results = client.search(query="latest AI developments")
print(results["results"])
```

### [POPULAR] Brave Search API
- URL: https://toolkiti.org/api/brave-search
- Description: Independent search index API with web, news, image, and video results. Privacy-focused.
- Pricing: free tier (2000 queries/mo), paid plans
- Authentication: API Key
- Rate Limit: 2000 queries/mo (free)
- Latency: <500ms
- Popularity: 88/100
- SDKs: curl, Python, JavaScript
- Endpoints:
  - https://api.search.brave.com/res/v1/web/search
- Key Features:
  - Independent search index
  - Web, news, image & video search
  - Privacy-focused, no tracking
  - Rich result data (ratings, reviews)
  - Free tier available
- Code Example (bash):
```bash
curl "https://api.search.brave.com/res/v1/web/search?q=AI+tools" \
  -H "Accept: application/json" \
  -H "Accept-Encoding: gzip" \
  -H "X-Subscription-Token: $BRAVE_API_KEY"
```

### [HOT] Serper API
- URL: https://toolkiti.org/api/serper
- Description: Fast Google Search API with rich SERP data including knowledge graph, shopping, and organic results.
- Pricing: free tier (2500 queries), pay-as-you-go from $50/mo
- Authentication: API Key
- Rate Limit: 2500 queries free once
- Latency: <500ms
- Popularity: 90/100
- SDKs: Python, Node.js, Go, curl
- Endpoints:
  - https://google.serper.dev/search
- Key Features:
  - Fast Google SERP API
  - Knowledge graph & shopping data
  - Organic, maps, images results
  - High reliability & uptime
  - Simple REST API
- Code Example (bash):
```bash
curl https://google.serper.dev/search \
  -H "X-API-KEY: $SERPER_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"q": "best AI tools 2026"}'
```

### [POPULAR] Jina AI API
- URL: https://toolkiti.org/api/jina-ai
- Description: Universal AI services: embeddings (8K+ dimensions), Reader API for web content, and segmentation.
- Pricing: free tier, pay-as-you-go
- Authentication: API Key
- Popularity: 80/100
- SDKs: Python, TypeScript, curl
- Endpoints:
  - https://api.jina.ai/v1/embeddings
  - https://r.jina.ai/ (Reader)
  - https://s.jina.ai/ (Segmenter)
- Key Features:
  - 8K+ dimensional embeddings
  - Reader API: web to LLM-ready content
  - Segmenter for text chunking
  - Open-source models available
  - CLIP & ColBERT architectures
- Code Example (bash):
```bash
curl https://api.jina.ai/v1/embeddings \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $JINA_API_KEY" \
  -d '{"input": ["Hello world"], "model": "jina-embeddings-v3"}'
```

### [POPULAR] Exa Search API
- URL: https://toolkiti.org/api/exa
- Description: AI-powered search engine built for embeddings-based semantic search. Finds content by meaning, not keywords.
- Pricing: free tier (1000 queries/mo), paid plans
- Authentication: API Key
- Rate Limit: 1000 queries/mo (free tier)
- Popularity: 82/100
- SDKs: Python, TypeScript, curl
- Endpoints:
  - https://api.exa.ai/search
  - https://api.exa.ai/contents
- Key Features:
  - Semantic search by meaning
  - Embeddings-based content discovery
  - Similarity & content retrieval
  - AI agent optimized
  - Developer-friendly API
- Code Example (bash):
```bash
curl https://api.exa.ai/search \
  -H "Content-Type: application/json" \
  -H "x-api-key: $EXA_API_KEY" \
  -d '{"query": "AI agents best practices", "numResults": 5}'
```

### [POPULAR] Firecrawl API
- URL: https://toolkiti.org/api/firecrawl
- Description: Web scraping and crawling API optimized for AI. Converts web pages into clean markdown or structured data.
- Pricing: free tier (500 pages/mo), paid plans
- Authentication: API Key
- Rate Limit: 500 pages/mo (free tier)
- Popularity: 77/100
- SDKs: Python, Node.js, Go, curl
- Endpoints:
  - https://api.firecrawl.dev/v1/scrape
  - https://api.firecrawl.dev/v1/crawl
- Key Features:
  - Web scraping for AI agents
  - Converts pages to clean Markdown
  - Supports JS-rendered pages
  - Crawl entire websites
  - Free tier available
- Code Example (bash):
```bash
curl https://api.firecrawl.dev/v1/scrape \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $FIRECRAWL_API_KEY" \
  -d '{"url": "https://example.com", "formats": ["markdown"]}'
```

---

## Image Generation (图像生成) — 4 APIs

###  Stability AI API
- URL: https://toolkiti.org/api/stability-ai
- Description: Generate images with Stable Diffusion 3, SDXL, and image-to-video models.
- Pricing: pay-as-you-go (per image)
- Authentication: API Key
- Popularity: 60/100
- SDKs: Python, TypeScript, curl
- Endpoints:
  - https://api.stability.ai/v1/generation/stable-diffusion-v3/text-to-image
- Key Features:
  - Stable Diffusion 3 & SDXL models
  - Text-to-image & image-to-video
  - ControlNet for precise control
  - Upscaling & outpainting tools
  - Commercial usage allowed
- Code Example (bash):
```bash
curl -X POST https://api.stability.ai/v1/generation/stable-diffusion-v3/text-to-image \
  -H "Authorization: Bearer $STABILITY_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "text_prompts": [{"text": "a cat wearing a hat"}],
    "samples": 1
  }'
```

### [POPULAR] Midjourney API
- URL: https://toolkiti.org/api/midjourney
- Description: Industry-leading AI image generation via Discord bot API. Produces artistic, photorealistic, and creative visuals.
- Pricing: subscription ($10-$60/mo), API via third-party
- Authentication: API Key, Discord Token
- Popularity: 75/100
- SDKs: Python (unofficial), curl
- Endpoints:
  - /imagine
  - /describe
  - /blend
- Key Features:
  - Industry-leading creative generation
  - Artistic & photorealistic output
  - Style consistency with references
  - Upscaling & variation tools
  - Active creative community
- Code Example (bash):
```bash
# Midjourney operates primarily through Discord
# Third-party APIs wrap the Discord bot:
# /imagine prompt: a serene mountain landscape --ar 16:9
```

###  Leonardo.AI API
- URL: https://toolkiti.org/api/leonardo-ai
- Description: AI image generation platform with fine-tuned models, real-time generation, and training API.
- Pricing: free tier (150 tokens/day), paid plans
- Authentication: API Key
- Popularity: 62/100
- SDKs: Python, curl
- Endpoints:
  - https://cloud.leonardo.ai/api/rest/v1/generations
- Key Features:
  - Fine-tuned generation models
  - Real-time canvas editing
  - Custom model training
  - API and web interface
  - Free daily tokens
- Code Example (bash):
```bash
curl https://cloud.leonardo.ai/api/rest/v1/generations \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $LEONARDO_API_KEY" \
  -d '{
    "prompt": "a futuristic cityscape",
    "modelId": "b24e16ff-06e3-43eb-8d33-4416c2d75876"
  }'
```

###  Ideogram API
- URL: https://toolkiti.org/api/ideogram
- Description: AI image generation with superior text rendering. Known for accurate text-in-image generation.
- Pricing: free tier, paid plans from $20/mo
- Authentication: API Key
- Popularity: 48/100
- SDKs: Python, curl
- Endpoints:
  - https://api.ideogram.ai/generate
  - https://api.ideogram.ai/upscale
- Key Features:
  - Superior text rendering in images
  - Accurate in-image typography
  - Style customization options
  - Upscaling API
  - Ideal for design & marketing
- Code Example (bash):
```bash
curl https://api.ideogram.ai/generate \
  -H "Content-Type: application/json" \
  -H "Api-Key: $IDEOGRAM_API_KEY" \
  -d '{
    "image_request": {
      "prompt": "A poster that says 'Hello World'",
      "aspect_ratio": "16/9"
    }
  }'
```

---

## Audio & Speech (音频语音) — 2 APIs

###  ElevenLabs API
- URL: https://toolkiti.org/api/elevenlabs
- Description: Text-to-speech, voice cloning, and sound effects generation with ultra-realistic voices.
- Pricing: free tier available, pay-as-you-go
- Authentication: API Key
- Popularity: 65/100
- SDKs: Python, Node.js, curl
- Endpoints:
  - https://api.elevenlabs.io/v1/text-to-speech
  - https://api.elevenlabs.io/v1/sound-generation
- Key Features:
  - Ultra-realistic text-to-speech
  - Voice cloning from samples
  - Sound effects generation
  - 29+ languages supported
  - Expressive & emotional speech
- Code Example (bash):
```bash
curl https://api.elevenlabs.io/v1/text-to-speech/21m00Tcm4TlvDq8ikWAM \
  -H "Content-Type: application/json" \
  -H "xi-api-key: $ELEVENLABS_API_KEY" \
  -d '{"text": "Hello from ElevenLabs!"}' \
  --output speech.mp3
```

###  AssemblyAI API
- URL: https://toolkiti.org/api/assemblyai
- Description: Speech-to-text, speaker diarization, sentiment analysis, and content moderation.
- Pricing: free tier available, pay-as-you-go
- Authentication: API Key
- Popularity: 50/100
- SDKs: Python, JavaScript, Go, Ruby, curl
- Endpoints:
  - https://api.assemblyai.com/v2/transcript
  - https://api.assemblyai.com/v2/stream
- Key Features:
  - High-accuracy speech-to-text
  - Speaker diarization
  - Sentiment analysis on transcripts
  - Content moderation
  - Real-time streaming support
- Code Example (bash):
```bash
curl https://api.assemblyai.com/v2/transcript \
  -H "Authorization: $ASSEMBLYAI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"audio_url": "https://example.com/audio.mp3"}'
```

---

## Vector Databases (向量数据库) — 3 APIs

### [POPULAR] Pinecone Vector Database
- URL: https://toolkiti.org/api/pinecone
- Description: Managed vector database for semantic search, RAG, and recommendation with millisecond latency.
- Pricing: free tier available, pay-as-you-go
- Authentication: API Key
- Popularity: 78/100
- SDKs: Python, Node.js, Go, Java, curl
- Endpoints:
  - https://{index}.svc.{environment}.pinecone.io/vectors/upsert
  - https://{index}.svc.{environment}.pinecone.io/query
- Key Features:
  - Managed vector database
  - Millisecond semantic search
  - RAG-optimized infrastructure
  - Serverless & pod-based indexes
  - Multi-cloud support
- Code Example (python):
```python
from pinecone import Pinecone
pc = Pinecone(api_key="$PINECONE_API_KEY")
index = pc.Index("my-index")
index.upsert(vectors=[
    {"id": "1", "values": [0.1, 0.2, 0.3]}
])
```

###  Weaviate Vector Database
- URL: https://toolkiti.org/api/weaviate
- Description: Open-source vector database with built-in vectorization, hybrid search, and GraphQL API.
- Pricing: free tier (cloud sandbox), pay-as-you-go, self-hosted
- Authentication: API Key
- Popularity: 64/100
- SDKs: Python, TypeScript, Go, Java, .NET
- Endpoints:
  - https://{cluster}.weaviate.cloud/v1/objects
  - https://{cluster}.weaviate.cloud/v1/graphql
- Key Features:
  - Open-source vector database
  - Built-in vectorization modules
  - Hybrid search (vector + keyword)
  - GraphQL API
  - Self-host or cloud
- Code Example (bash):
```bash
curl https://your-cluster.weaviate.cloud/v1/graphql \
  -H "Authorization: Bearer $WEAVIATE_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"query": "{ Get { Document(nearText: {concepts: ["AI"]}) { title } } }"}'
```

###  Qdrant Vector Database
- URL: https://toolkiti.org/api/qdrant
- Description: High-performance vector search engine written in Rust, for semantic search and recommendations.
- Pricing: free tier (cloud), pay-as-you-go, self-hosted (free)
- Authentication: API Key
- Popularity: 63/100
- SDKs: Python, TypeScript, Go, Rust, Java
- Endpoints:
  - https://{cluster}.cloud.qdrant.io:6333/collections
  - https://{cluster}.cloud.qdrant.io:6333/collections/{name}/points/search
- Key Features:
  - High-performance (Rust-based)
  - Rich filtering & payload support
  - Scalable distributed architecture
  - Self-hosted (free) & cloud tiers
  - gRPC & REST interfaces
- Code Example (bash):
```bash
curl -X POST https://your-cluster.cloud.qdrant.io:6333/collections/my-collection/points/search \
  -H "Content-Type: application/json" \
  -H "api-key: $QDRANT_API_KEY" \
  -d '{"vector": [0.1, 0.2, 0.3], "limit": 5}'
```

---

## GPU Cloud (GPU 云) — 2 APIs

###  Modal GPU Cloud
- URL: https://toolkiti.org/api/modal
- Description: Serverless GPU compute platform for AI inference, batch processing, and model deployment.
- Pricing: pay-as-you-go (per second), free tier available
- Authentication: API Key (Token)
- Popularity: 52/100
- SDKs: Python SDK
- Endpoints:
  - https://api.modal.com/v1/functions
  - https://api.modal.com/v1/volumes
- Key Features:
  - Serverless GPU compute
  - Pay-per-second pricing
  - Python SDK with local dev
  - Auto-scaling infrastructure
  - Volumes & secret management
- Code Example (python):
```python
import modal
app = modal.App("my-app")

@app.function(gpu="A100")
def generate(prompt: str):
    # Your AI inference code here
    return f"Generated from: {prompt}"
```

###  RunPod GPU Cloud
- URL: https://toolkiti.org/api/runpod
- Description: GPU cloud platform with serverless endpoints, pods, and auto-scaling for AI workloads.
- Pricing: pay-as-you-go (per hour/sec), cheap spot instances
- Authentication: API Key
- Popularity: 50/100
- SDKs: Python, curl
- Endpoints:
  - https://api.runpod.ai/v2/{endpoint_id}/runsync
  - https://api.runpod.ai/v1/pods
- Key Features:
  - GPU pods & serverless endpoints
  - Cheap spot instances
  - Auto-scaling for workloads
  - Pre-built templates available
  - Community serverless templates
- Code Example (bash):
```bash
curl https://api.runpod.ai/v2/YOUR_ENDPOINT_ID/runsync \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $RUNPOD_API_KEY" \
  -d '{"input": {"prompt": "Hello"}}'
```

---

## Developer Tools (开发工具) — 4 APIs

###  LangChain Framework
- URL: https://toolkiti.org/api/langchain
- Description: Framework for building LLM-powered applications with chains, agents, RAG, and tool integration.
- Pricing: open-source (free), LangSmith/LangGraph Cloud have paid tiers
- Authentication: API Key (for cloud services)
- Popularity: 55/100
- SDKs: Python, TypeScript/JavaScript, Java (LangChain4j)
- Endpoints:
  - N/A (Framework, not API)
- Key Features:
  - Chain composition framework
  - Agent & tool integration
  - RAG building blocks
  - LangSmith observability
  - Multi-provider support
- Code Example (python):
```python
from langchain_core.prompts import ChatPromptTemplate
from langchain_openai import ChatOpenAI

prompt = ChatPromptTemplate.from_template("Tell me a {topic} joke")
model = ChatOpenAI(model="gpt-4o")
chain = prompt | model

result = chain.invoke({"topic": "programming"})
print(result.content)
```

### [HOT] GitHub REST & GraphQL API
- URL: https://toolkiti.org/api/github-api
- Description: Full GitHub platform API: repos, issues, PRs, Actions, search, code scanning, and more.
- Pricing: free (5000 req/hr for authenticated), unlimited for public repos
- Authentication: Personal Access Token, OAuth App, GitHub App
- Rate Limit: 5000 req/hr (authenticated)
- Popularity: 92/100
- SDKs: Octokit (JS/TS), PyGithub, Go-Github, curl
- Endpoints:
  - https://api.github.com
  - https://api.github.com/graphql
- Key Features:
  - Full GitHub platform API
  - REST & GraphQL interfaces
  - Repos, issues, PRs, Actions
  - Code scanning & security
  - Generous free tier
- Code Example (bash):
```bash
curl -H "Authorization: token $GITHUB_TOKEN" \
  -H "Accept: application/vnd.github+json" \
  https://api.github.com/user/repos?per_page=5
```

### [POPULAR] Supabase API
- URL: https://toolkiti.org/api/supabase
- Description: Open-source Firebase alternative: PostgreSQL database, auth, realtime subscriptions, and storage API.
- Pricing: free tier (2 projects, 500 MB DB), paid plans
- Authentication: API Key (anon/service_role), OAuth
- Popularity: 76/100
- SDKs: JavaScript/TypeScript, Python, Dart, Go, Swift, Kotlin
- Endpoints:
  - https://[project].supabase.co/rest/v1/
  - https://[project].supabase.co/auth/v1/
- Key Features:
  - Open-source Firebase alternative
  - PostgreSQL database with REST API
  - Built-in auth & Row Level Security
  - Realtime subscriptions
  - File storage
- Code Example (bash):
```bash
curl "https://your-project.supabase.co/rest/v1/users" \
  -H "apikey: $SUPABASE_ANON_KEY" \
  -H "Authorization: Bearer $SUPABASE_ANON_KEY"
```

###  Resend API
- URL: https://toolkiti.org/api/resend
- Description: Email API for developers. Send transactional and marketing emails with high deliverability.
- Pricing: free tier (3000 emails/mo for first 3 months), paid plans
- Authentication: API Key
- Popularity: 55/100
- SDKs: JavaScript/TypeScript, Python, Go, PHP, Ruby, Elixir, curl
- Endpoints:
  - https://api.resend.com/emails
- Key Features:
  - Email API for developers
  - High deliverability rates
  - React Email templates
  - Analytics & tracking
  - Generous free tier
- Code Example (bash):
```bash
curl https://api.resend.com/emails \
  -H "Authorization: Bearer $RESEND_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "from": "onboarding@resend.dev",
    "to": ["user@example.com"],
    "subject": "Hello!",
    "text": "Welcome to Resend"
  }'
```

---

## Machine Learning (机器学习) — 2 APIs

###  Replicate API
- URL: https://toolkiti.org/api/replicate
- Description: Run open-source ML models in the cloud. Supports Llama, Stable Diffusion, Whisper, and thousands more.
- Pricing: pay-as-you-go (per prediction)
- Authentication: API Key
- Popularity: 73/100
- SDKs: Python, Node.js, curl
- Endpoints:
  - https://api.replicate.com/v1/models
  - https://api.replicate.com/v1/predictions
- Key Features:
  - Thousands of open-source ML models
  - Llama, Stable Diffusion, Whisper
  - Serverless API with autoscaling
  - Custom model deployment
  - Pay-per-prediction pricing
- Code Example (python):
```python
import replicate
output = replicate.run(
    "meta/meta-llama-3.1-405b-instruct",
    input={"prompt": "Hello!"}
)
for chunk in output:
    print(chunk, end="")
```

###  Hugging Face Inference API
- URL: https://toolkiti.org/api/huggingface
- Description: Access 200,000+ models for inference including transformers, diffusion, and audio models.
- Pricing: free tier, pay-as-you-go for accelerated inference
- Authentication: API Key (Bearer Token)
- Popularity: 50/100
- SDKs: Python, JavaScript, curl
- Endpoints:
  - https://api-inference.huggingface.co/models/{model_id}
- Key Features:
  - 200,000+ models for inference
  - Transformers, diffusion, audio
  - Free inference API
  - Accelerated inference available
  - Community-driven model hub
- Code Example (bash):
```bash
curl https://api-inference.huggingface.co/models/meta-llama/Meta-Llama-3.1-405B-Instruct \
  -H "Authorization: Bearer $HF_API_KEY" \
  -d '{"inputs": "Hello!", "parameters": {"max_new_tokens": 100}}'
```

---

## Site Information
- Site: ToolKiti - Structured API References for AI Agents
- URL: https://toolkiti.org
- Full JSON data: https://toolkiti.org/api/data
- API comparison: https://toolkiti.org/compare
- Text format: https://toolkiti.org/api/ai.txt
- Bilingual: English (primary), Chinese
- Search available: On homepage