AkashML Inference API

name: AkashML Inference API
description: OpenAI-compatible AI inference API providing access to open-source LLMs hosted on decentralized
  compute across 80+ global datacenters. Supports chat completions with models including Llama 3.3-70B,
  DeepSeek V3, Qwen3-30B-A3B, and others. Drop-in replacement for OpenAI API clients. New accounts receive
  $100 in free AI token credits.
image: https://akash.network/favicon.ico
humanURL: https://akash.network/blog/akashml-managed-ai-inference-on-the-decentralized-supercloud/
baseURL: https://api.akashml.com
tags:
- AI
- LLM
- Inference
- OpenAI Compatible
- Chat Completions
- Llama
- DeepSeek
properties:
- type: Documentation
  url: https://akashml.com/
- type: Playground
  url: https://playground.akashml.com
contact:
- FN: AkashML Support
  url: https://akash.network/community
X-pricing:
- plans:
  - name: Pay-Per-Token
    description: Pay-per-token pricing with no subscription. New accounts receive $100 in free credits.
      Pricing varies by model.
    url: https://akashml.com/
    price: 0.13
    unit: per million input tokens (Llama 3.3-70B)
X-rateLimit:
- rate:
  - requests: 0
    unit: none
    notes: No formal rate limits published. Service advertises no capacity caps. Sub-200ms response times
      across 80+ global datacenters.
X-authentication:
- type: Bearer
  description: Bearer token authentication via Authorization header. Obtain API key from akashml.com account
    dashboard.
  url: https://akashml.com/