What MetaDOS does

MetaDOS LLC, doing business as DOS AI, is a Delaware-incorporated private AI infrastructure company headquartered in Wilmington, DE with an operational office in Singapore and a remote-first team distributed across 10+ countries. The company traces its origins to DOS Labs, a Web3 gaming and blockchain business that developed the MetaDOS battle royale game on Avalanche's DOS Chain and closed a $2.45 million financing round in April 2024 with AZA Ventures, Avalon Wealth Club, Brainworks Ventures, GD10 Ventures, and Kangaroo Capital. The organization subsequently pivoted to AI infrastructure, relaunching itself as DOS AI and building a platform-plus-modules stack on top of an OpenAI-compatible inference API at api.dos.ai/v1. The core inference layer runs self-hosted open-source models (notably Qwen3.5-35B-A3B, Llama 4, DeepSeek V3, Mistral) on dedicated RTX Pro 6000 GPUs in Asia-Southeast 1 via vLLM, with a proprietary 15-dimension smart router (dos-auto) that selects the optimal model per request based on cost, latency, and complexity. The 41-model catalog also includes embeddings (Qwen3-Embedding-4B, 2560-dim), image generation (FLUX.1, SDXL), video (Wan 2.6/2.7), music (minimax-music-2.5+), and dedicated code models, with response caching and structured output support.

Built on this inference core, DOS AI offers three modules: DOSClaw, a managed hosting platform for the open-source OpenClaw agent runtime (Apache 2.0) that deploys autonomous AI agents to 22+ messaging channels (Telegram, Slack, Discord, WhatsApp, Messenger, Line, Zalo, WeChat) at subscription tiers of $5–$20/month; DOSafe, a safety and threat-intelligence engine aggregating 3.93M+ entries from 19+ scrapers, with AI text/image/video/audio detection (AUROC 0.88), entity risk scoring, and face/voice verification; and a GPU Cloud offering per-second RTX Pro 6000 Blackwell rental starting at $0.99/hour in Southeast Asia. The company is SOC2 Type II compliant and GDPR/PDPA ready, with 10K+ verified DOS.Me identities feeding risk signals across the stack. Revenue is generated through a multi-pronged model: usage-based token pricing, freemium credits ($5–$10 per new user), tiered subscriptions, GPU rental, and partner API licensing — distributed via a four-channel GTM (self-serve PLG, API-first developer adoption, enterprise field sales, and community-led growth via Discord, Telegram, and GitHub).

The broader DOS ecosystem also retains Web3 products from the company's origins: the MetaDOS game, DOS Chain (Avalanche-based L1), DOScan (block explorer), DOSwap (swap), DOS.Me (identity), and OverMint (NFT). Leadership is drawn primarily from Vietnam's gaming and blockchain sector — CEO Quang Nguyen (formerly of VNG, Appota, Gameloft, Zen Group), COO Anh Le (formerly CPO at Heroes & Empires, Producer at Gameloft), and CTO Minh Pham (formerly of Heallios, Wolffun) — reflecting the company's regional roots and operational footprint in Southeast Asia.

Name

MetaDOS

Legal name

MetaDOS LLC

Website

https://dos.ai

Company type

Private

Founded year

2024

Operating status

Operating

Headcount range

11–50 employees

Short description

MetaDOS LLC (DOS AI) provides an OpenAI-compatible AI inference API on self-hosted open-source models, plus managed AI agent hosting (DOSClaw), safety/threat intelligence (DOSafe), and on-demand GPU cloud in Southeast Asia, targeting developers, startups, and SMBs.

Ownership category

akta.pro rank

Product category

AI Infrastructure / LLM Inference Platform

NAICS

Software Publishers (5132), Computer Systems Design and Related Services (54151), Custom Computer Programming Services (541511)

SIC

Services-Computer Programming Services (7371), Services-Computer Programming, Data Processing, Etc. (7370)

akta.pro primary industry

Model Deployment, Serving & Inference Platforms (HDAAABAF)

akta.pro secondary industries

Model Serving, Inference & Deployment Platforms (APIs, Edge/On-Prem) (HDAEANAC), End-to-End Enterprise AI Platforms (MLOps & Model Lifecycle Management) (HDAEANAA)

Keywords

• AI inference API
• LLM hosting platform
• Managed AI agents
• GPU cloud rental
• Threat intelligence API

Headquarters

HQ city

Wilmington

HQ country

United States

HQ region

North America

Offices2 records

Markets served

GTM type

B2B

Offering type

Software

Cost components

Technology or R&D, Infrastructure, Personnel, Operations, Marketing or Sales

Revenue model

1. Pay-as-you-go LLM Inference: Token-based usage pricing per 1M tokens (input and output) across the model catalog (Qwen, Llama, DeepSeek, Mistral, etc.). New accounts receive $5.00 in free credits; minimum top-up is $5.00; billed monthly via Stripe.
2. DOSClaw Subscription Tiers: Tiered subscriptions: Free 7-day trial ($5 AI credits), Plus at $5/mo (1 standard instance + $5 AI credits/month), Pro at $20/mo (2 plus instances + $20 AI credits/month). Plus extras available at standard DOS.AI rates.
3. GPU Cloud Rental: Per-second billing for on-demand GPU instances (RTX Pro 6000 Blackwell) in Southeast Asia, ranging from $0.99/hr entry tier up to $9.29/hr for 8x GPU configurations; reserved instances save 25%.
4. DOSafe Partner API: API-key-based access (X-Api-Key) for partner integrations covering entity/URL safety checks, bulk checks, AI text/image/video/audio detection, and face/voice verification. Enterprise/partner pricing by negotiation.
5. Freemium Free Credits: $5–$10 in free credits for new users on signup (no credit card required) to drive developer adoption of the inference API and GPU cloud; free credits do not expire.
6. Media Generation Billing: Video and audio billed per second of generated media (not per token); $10/1000 seconds for Wan 2.7 video, $5/1000 seconds for MiniMax Music 2.5+ audio.

Pricing tiers

Go-to-market motion3 records

Distribution channels7 records

Marketing channels8 records

Core offering

MetaDOS (dba DOS AI) operates an OpenAI-compatible AI inference API platform that serves open-source and proprietary large language models from dedicated RTX Pro 6000 GPUs in Southeast Asia, with pay-as-you-go token-based pricing, smart routing (dos-auto), and response caching. Around this core inference API, it bundles managed AI agent hosting (DOSClaw, deployable to 22+ messaging channels), a unified safety and threat intelligence API (DOSafe with 3.93M+ entries), an on-demand GPU cloud, and a Web3 ecosystem of sub-brands (DOS Chain, DOS.Me, DOScan, DOSwap, OverMint, MetaDOS game).

Product overview

MetaDOS LLC (doing business as DOS AI) operates a unified AI infrastructure platform with a platform-plus-modules architecture. The core product is DOS.AI, an OpenAI-compatible inference API providing access to open-source LLMs (Llama, Qwen, DeepSeek, Mistral, Gemma), embeddings (Qwen3-Embedding-4B), and media generation (Wan video, music). Built on top of this core, DOSClaw adds managed hosting for autonomous AI agents deployable to 22+ messaging channels, and DOSafe adds a unified safety/threat-intelligence gateway with AI text/image/video/audio detection and entity risk scoring. The company also operates a GPU Cloud (per-second RTX Pro 6000 Blackwell rentals in Southeast Asia) and a broader DOS Ecosystem of Web3 products including the MetaDOS battle royale game, DOS Chain (Avalanche-based L1), DOS.Me (identity), DOScan (block explorer), DOSwap (swap), and OverMint (NFT).

Differentiator

Problem solved

Functional benefit

Brands

• DOS AI: The primary AI inference API platform offering OpenAI-compatible access to leading open-source language models, embeddings, image, video and audio generation on dedicated GPU infrastructure.
• DOSClaw
• DOSafe

Products and services

• DOS.AI Inference API Platform OpenAI-compatible AI inference API platform providing access to 41 leading open-source and proprietary models (Llama, Qwen, DeepSeek, Mistral, Gemma, etc.) for chat, code, vision, embeddings, image, video, and audio. Features smart routing (dos-auto), pay-as-you-go token pricing, response caching, and SOC2 Type II / GDPR / PDPA compliance. Targeted at developers, startups, mid-market tech companies, and enterprise customers.
• DOSClaw AI Agent Platform Managed hosting platform for the open-source OpenClaw AI agent runtime, enabling deployment of autonomous AI agents (Sales Bot, Customer Service, Personal Assistant, Content Creator) to 22+ messaging channels. Includes built-in AI credits, knowledge base (RAG), templates, plugin ecosystem with 50+ skills, and a 15-layer security stack with per-agent container isolation. Plans: Free 7-day trial, Plus $5/mo, Pro $20/mo.
• DOSafe Partner API (Safety & Threat Intelligence) Unified safety and threat intelligence API aggregating 3.93M+ entries from 19 scrapers plus on-chain EAS attestations. Endpoints for entity safety check, bulk check, URL check, AI text/image/video/audio detection (AUROC 0.88 for audio via BEATs + mHuBERT), and face/voice verification. Used by partners such as the Bexly wallet for transaction safety.
• DOS GPU Cloud On-demand GPU rental service with per-second billing and 1x/2x/4x/8x RTX Pro 6000 Blackwell configurations in the Southeast Asia region, starting at $0.99/hr. Supports model training, inference/serving, and R&D with full root access, custom Docker images, persistent storage, and pre-built PyTorch/TensorFlow/Docker templates.
• MetaDOS (Web3 Battle Royale Game) Web3 battle royale game built on the Avalanche-based DOS Chain, developed by DOS Labs. Initial version released on Steam with full release scheduled for May 2024. Provides players ownership of in-game assets via blockchain.

Quantifiable outcome

• 67% faster than leading alternatives in request latency
• +6 more outcomes

Named customers6 records

Ideal customer profiles4 records

Technology focussed Yes

API detail

Has API

Yes

API docs

API detail

Core technology

AI maturity

App detail

Integration12 records

AI capability17 records

Feature8 records

Partnerships

Five partnerships are on record, tiered core and minor.

• OpenClaw (open-source AI agent runtime)coreStrategic or Co-development PartnerDOSClaw is the managed-service hosting platform built on top of the open-source OpenClaw runtime (Apache 2.0, github.com/nicejoy/OpenClaw). DOS AI handles servers, updates, and scaling so users can deploy agents without DevOps. OpenClaw is co-developed/maintained by the MetaDOS team under the github.com/nicejoy organisation.
• StripecoreTechnology or IntegrationStripe is the payment processor for credit purchases and subscription billing; payments are non-refundable except as required by law or at DOS AI's sole discretion, and Stripe is referenced in both Terms of Service and Privacy Policy.
• DeepInfracoreTechnology or IntegrationDeepInfra is the partner cloud provider serving Meta's Llama 4 Maverick and Llama 4 Scout models on the DOS AI catalog — listed as 'Meta / DeepInfra' in the Available Models reference.
• Telegram / Slack / Discord / WhatsApp / Messenger / Line / Zalo / WeChat (messaging platforms)coreChannel Partner/ Reseller/ DistributorDOSClaw agents integrate natively with 22+ messaging platforms via user-provided bot tokens, enabling distribution of DOS-built agents across major chat ecosystems.
• Bexly (example DOSafe integration customer)minorImplementation/ SI/ Consulting PartnerBexly is cited in DOSafe documentation as an example integrator using the /check endpoint to evaluate wallet recipients before transactions — illustrative of how partner apps consume the safety API.

Scale indicators9 records

Recent moves6 records

Expansion highlights6 records

Direct peers

• Together AI: Together AI offers an OpenAI-compatible inference API for open-source LLMs with similar self-hosted GPU infrastructure and token-based pricing. Most direct head-to-head competitor to DOS in the open-model inference layer.
• Fireworks AI: Fireworks AI provides low-latency OpenAI-compatible inference and fine-tuning for open-source models on dedicated GPUs — direct overlap with DOS on inference pricing, latency, and drop-in compatibility.
• Groq: Groq is a fast LLM inference provider competing on latency (its core differentiator) with custom LPU hardware. DOS markets '67% faster than leading alternatives' on the homepage — Groq is one of the alternatives most likely in that comparison.
• DeepInfra: DeepInfra is both a partner (serving Llama 4 Maverick/Scout on the DOS catalog) and a direct competitor offering OpenAI-compatible inference for open-source LLMs at similar price points.
• Replicate: Replicate runs a serverless inference API for open-source ML models (LLMs, image, audio, video) with usage-based pricing — directly comparable to DOS's pay-as-you-go inference across multimodal endpoints.
• Anyscale: Anyscale (built on Ray) offers OpenAI-compatible LLM serving and managed compute for AI workloads, comparable to DOS's combined inference and GPU cloud offering.
• RunPod: RunPod provides on-demand GPU cloud rentals (similar H100/consumer-grade SKUs) with per-second billing — closely matches DOS GPU Cloud's $0.99/hr entry pricing and developer self-serve motion.
• Lambda Labs: Lambda provides GPU cloud plus hosted LLM inference for open-source models — comparable in combining raw GPU rental with model-serving APIs, including SOC2-ready enterprise options.

Broad incumbents

• OpenAI: OpenAI is the dominant incumbent whose API compatibility DOS explicitly targets as a switching destination. DOS's value prop is framed as 'drop-in replacement for OpenAI' with lower cost.
• Hugging Face: Hugging Face operates Inference Endpoints, serverless inference, and Spaces — comparable infrastructure and open-source-model distribution surface that overlaps with DOS's developer-led offering.

Market position

Strengths5 records

Weaknesses5 records

Competitive moat6 records

Key risks7 records

Key highlights6 records

Customer concentration

• GitHubgithub.com/DOS/status
• LinkedInlinkedin.com/company/DOS
• Xx.com/MetaDOS
• Facebookfacebook.com/RunDOS

Compliance4 records

Revenue estimate

Valuation estimate

Number of profiles

Profiles7 records

Subsidiaries1 record

Funding overview

Funding rounds4 records

Investors6 records

Funding detail is available on the Subscription and Enterprise plan.Contact sales →

M&A

Investments

M&A and investment is available on the Subscription and Enterprise plan.Contact sales →

Frequently asked questions about MetaDOS