Official Platforms
50Official or cloud vendor API platforms, most authoritative and stable
OpenAI official platform providing GPT series model APIs
Google AI Studio, official Gemini API platform
Claude official API platform, providing Claude series models
DeepSeek official API platform, Input: ¥4/M Tokens, Output: ¥16/M Tokens
Moonshot AI Kimi, ultra-long context support, excels at long text processing
Alibaba Tongyi Qianwen official API
Baidu Wenxin Yiyan official API
Volcano Engine Ark, large model service platform
Zhipu AI GLM BigModel, providing GLM-4 and other models
Mistral AI official platform, providing Mistral series models
Cohere official API platform
Groq Cloud, ultra-fast LLM inference
GitHub Models, Azure-hosted model APIs
Hugging Face Inference API, aggregating multiple open-source models
Alibaba ModelScope, aggregating Qwen / DeepSeek and other open-source models
Cerebras Inference, free/quota-based LLM inference by token
InfiniAI GenStudio, featuring DeepSeek R1/V3 full-power free tokens
xAI Grok, designed with humor and real-time information access
Xiaomi MiMo reasoning model platform, providing official MiMo series APIs
MiniMax by Xiyu Technology, leading multimodal capabilities, top-tier voice synthesis and video generation
iFlytek Spark, strong Chinese understanding, fast response, unlimited total tokens
StepFun, focused on multimodal large model research and development
01.AI by Kai-Fu Lee, active Yi series model open-source ecosystem
Baichuan AI, Chinese large model optimized for medical and financial scenarios
SenseTime SenseNova, CV background, strong multimodal capabilities
Perplexity AI, conversational search engine with real-time web access
Midjourney, world-leading AI image generation platform
Stability AI, official Stable Diffusion series models
Runway ML, leader in AI video generation and editing tools
ElevenLabs, industry-leading AI voice synthesis and cloning platform
Character.AI, open platform for creating and chatting with AI characters
Jasper AI, enterprise AI writing assistant for marketing content generation
Copy.ai, AI copywriting tool for quick marketing content creation
Synthesia, AI video generation platform with digital avatars
HeyGen, AI digital human video generation with multilingual lip-sync
Replicate, open-source AI model running platform
Together AI, open-source model inference acceleration platform
Fireworks AI, fast inference platform supporting mainstream open-source models
AI21 Labs, Jurassic-3 series large models, enterprise-grade NLP solutions
Aleph Alpha, European open-source large model with multilingual support
Databricks, DBRX open-source model, enterprise AI platform
NVIDIA NIM, GPU giant model inference microservices
ByteDance Doubao large model, strong voice synthesis and visual understanding
Tencent Hunyuan large model, outstanding Chinese creation and logical reasoning
360 AI Brain, enterprise-grade large model platform by 360
NetEase Fuxi, leading game AI, virtual humans and animation generation
ModelBest MiniCPM, leading edge-side small model performance
Kunlun Tech Tiangong AI, search-enhanced large model
Chumenwenwen Sequence Monkey, voice interaction and smart hardware
Langboat Mengzi large model, optimized for finance and marketing
Cloud Providers
19Large model API services provided by major cloud vendors, enterprise-grade stability
Alibaba Cloud Bailian, most domestic models, 257+
Tencent Cloud TI-ONE training platform, Hunyuan + open-source model deployment
Huawei Cloud Pangu, high government/enterprise compatibility, compliance advantages
Baidu Qianfan platform, ERNIE series + third-party models
AWS Bedrock, 30+ models, enterprise governance, requires overseas account
Microsoft Azure AI, 20+ models, ISO/SOC certified, preferred by multinationals
Qiniu AI, 50+ models, domestic direct connection, compatible with OpenAI/Anthropic
UCloud Model Square, limited open-source models, fine-tuning deployment
Tianyi Cloud large model service, telecom operator background
Mobile Cloud large model, telecom cloud, government and SOE scenarios
Cloudflare Workers AI, 10K free requests daily, edge computing deployment
National Supercomputing Internet Platform, supercomputing and model API services
JD Cloud large model service, optimized for e-commerce and logistics
Unicom Cloud large model service, telecom operator background
Moore Threads GPU cloud service, domestic AI computing platform
Infini-AI, large model computing power service platform
Compshare GPU cloud servers, AI training and inference computing rental
Jiuzhang AI Cloud, focused on Claude Code programming assistant packages
GitCode AI, AI programming assistant by open-source code hosting platform
Third-Party Relay
35Third-party API relay services, more cost-effective than official
Widest global model coverage, flexible routing, 400+ models
Domestic open-source model inference acceleration, DeepSeek/Qwen optimized, latency as low as 30ms
Enterprise AI model API relay, 400+ models, 300K+ monthly visitors
AI API aggregation relay platform, multi-model unified calling, stable domestic connection
Pay-per-use enterprise AI resource platform, latest domestic and overseas model APIs
Community-recognized serious relay, excellent access experience
US company background, cloud service provider authorized, full models near official price
Triple protocol native support, 480+ models, 99.99% SLA, enterprise preferred
50+ models, triple protocol compatible, TTFB 310ms, extremely low latency
100-200+ models, fast new model listing, multi-channel backup
Wide model coverage, high stability, enterprise relay service
Emerging domestic aggregation platform, outstanding cost performance
Veteran API station, supports many models
First-generation CC/CX/Gemini programming tool relay, official pure channels
Most stable official enterprise channel, 36 months stable operation
Enterprise large model API aggregation platform, one Key covers all models
Focus on official APIs, 99.9% availability, 24/7 operation
Alibaba team background, top Alibaba Cloud agent, Qwen official direct 50% off
Low-price Claude channel, great for coding
Stable for months, currently all Claude models at 1.9 discount
Ultra cost-effective, rate 0.1, source supply for Claude GPT Gemini
Super cheap codex, agent rate as low as 0.025
Focus on GPT models, full power, pay-per-use 0.2 rate
Simple Claude relay station, from private Claude account pool
Programming dedicated API relay platform, stable model access
Self-built server nodes, many small-parameter open-source models
Not just AI APIs, also supports YouTube, Twitter, financial data APIs
Besides text models, also comprehensive video and image models
Supports many models, covers Claude/GPT/Gemini, fast new model updates
Focus on Alibaba series models, pursue long-term stability and low price
Beginner and personal users have super cost-effective packages
Popular domestic relay, attractive discount but mainly stable for Claude
Highly recommended, active on Twitter, has own tech development
OpenCode, overseas AI programming assistant aggregation, 14+ models, from $10/mo
ZenMux AI, overseas aggregation platform, supports multi-model unified calling
Open Source Tools
18Open source projects and tools for building relay stations
Next-gen AI model aggregation management system based on One API, 7.5K+ GitHub Stars
LLM API management & distribution system, 24.2K+ GitHub Stars, widely used as relay foundation
One-click free deployment of your private ChatGPT web app, 55.6K+ GitHub Stars
Open-source high-performance chatbot framework, voice synthesis, multimodal, plugin system
Powerful relay platform based on One-API fork, enhanced admin features
Open-source browser plugin, aggregate management for One-API and New-API accounts
Claude Code relay server, key management, usage statistics and forwarding
Claude relay management tool (CLI+GUI), one-click switch, export environment variables
Open-source LLM gateway, 100+ models unified calling, smart routing and cost tracking
Subscription to API service, supports multi-platform subscription account conversion to unified API
Command-line proxy API tool, supports multi-protocol forwarding and request conversion
Lightweight API relay service, supports request forwarding, rate limiting and logging
LLM API proxy server, supports model routing and request caching
OpenAI API proxy service, supports multi-key polling and request rate limiting
AI API aggregation proxy, supports automatic failover for multiple upstream services
ChatGPT API proxy service, supports WebSocket and streaming responses
One API multi-model management system, supports token management and usage statistics
GPT API proxy, supports load balancing and request retry
Clients
6API client tools
Supports Windows, Android, macOS, web, free, compatible with various APIs
Desktop-focused, compatible with various APIs, beautiful and easy-to-use interface
Online open-source WebUI, supports various API interfaces
iPad dedicated local client, supports various APIs
AI client designed specifically for macOS
Rapidly growing, cool interface, open-source and free
Benchmarks
8AI evaluation and benchmarking platforms
Hugging Face Open LLM Leaderboard, tracking and evaluating leading open-source models
Human preference-based anonymous chatbot battles and rankings, Elo rating system
Anti-contamination dynamic evaluation, monthly updated question bank, 6 categories
Leading domestic open-source large model evaluation system
Chinese general large model comprehensive benchmark, 100% question replacement every 2 months
Daily updated AI model comprehensive ranking, comparing performance, price and context window
Compare 100+ AI models intelligence, price, performance, speed and context window
LLM ranking focused on code editing ability, evaluating actual programming task performance