1import OpenAI from openai23client = OpenAI(4 base_url = "https://api.lumesh.ai/openai/v1",5 api_key = "sk-xxx",6)7chat_completion_res = client.chat.completions.create(8 model = "llava:7b",9 messages = [10 {11 "role": "system",12 "content": "You are a professional AI documentation assistant.",13 },14 {15 "role": "user",16 "content": "Write me a 4000-word story.",17 }18 ],19 stream = True,20 temperature = 2,21 max_tokens = 100022)23for chunk in chat_completion_res:24 print(chunk.choices[0].delta.content or "", end="")25
Simple to use, easy to integrate
Large model API services provide easy-to-integrate multimodal model APIs for various AI applications, covering large language models, images, audio, and video, helping you easily build and scale intelligent solutions.
Functions
Decentralized AI Cluster
Integrate global computing resources to
build an AI cluster based on DePIN.
Deliver Ubiquitous AI Inference
Services to Users
All types of devices, including smartphones, laptops, and personal computers, can easily access and use efficient AI model inference services.
Provide AI Development Support
for Developers
Offer OpenAI-compatible APIs to supply developers with
online AI inference services like Llama and Gemma,
accelerating the development process.
Provide Cost-effective AI Computing
Resources for Enterprises
Through Lumesh, enterprises can obtain AI
computing resources with higher cost-
effectiveness than traditional cloud computing
service providers.
Features
Always Online
Complete decentralization ensures that Lumesh is always online.
Ultra-Fast Response
The nearest and best-matched worker is
scheduled to complete AI inference tasks,
obtaining extremely fast response speeds
through streaming interfaces.
Innovative PoW
Workers prove their work through AI
tasks and are rewarded according to the
work they complete.
Earn More
Workers can earn rewards for completing AI inference tasks or for being
effectively online; users can also earn rewards for using AI services.
Dynamic Supply and Demand Balance
Through Lumesh's API, developers can create a diverse range of
AI applications, achieving a balanced network of supply and
demand, and continuously optimizing the user experience.
Network Metrics
0
Total Q&As for User
0
Total AI Inference tasks
0
Online Workers
0
Total U Points
0
Total W Points
Roadmap
Phase 1: Testnet Development
Website development
Whitepaper creation
Container Cloud development
Worker client development
Worker dashboard development
Ask AI to earn App development
Phase 2: Testnet Launch and Alpha Season
Release and launch of Lumesh testnet v0.1
Test and optimization of validator, worker client, worker dashboard and ask to earn AI App
Target number of users: 1000+
Target number of workers: 100+
Phase 3: Testnet Season 1~4
OpenAI-compatible APIs
Container Cloud Service
New UI design implementation
Release of Lumesh testnet v0.9
Global marketing initiatives
Target number of AI applications: 5+
Target number of users: 10,000+
Target number of workers: 5,000+
Target number of validators: 50+
Phase 4：Mainnet Beta Launch and Season 5 to 8
Release of Lumesh mainnet beta v1.0
TGE
Mainnet performance optimization
Target number of AI applications: 50+
Target number of users: 100,000+
Target number of workers: 50,000+
Target number of validators: 500+
Phase 5: Mainnet Beta Optimization and Season 9 to 10
Continuous optimization of the mainnet beta
Target number of AI applications: 500+
Target number of users: 1 million+
Target number of workers: 100,000+
Target number of validators: 1,000+
