Back to News
RadixArkAugust 19, 2026

Miles

Open Source
Reinforcement Learning
LLM Post-training
Agentic AI
Distributed Systems

Explore Miles

Visit the official website to learn more and get started

### TL;DR

Miles is an open-source reinforcement learning (RL) framework designed for large-scale model post-training, including agentic workloads. It integrates SGLang for high-throughput rollouts with Megatron-LM or FSDP2 for distributed training, providing a stable and efficient stack for frontier models.

Key Insights & Metrics

Pricing
Free and Open-source
Cost structure
Version
0.1
Current release version
Hardware
NVIDIA Blackwell, Hopper, and Ampere GPUs; AMD Instinct GPUs via ROCm
Compute requirements
Category
Open Source
Licensing model
Region
United States
Primary region

Key Features

  • Native SGLang rollout with fully asynchronous scheduling
  • Rollout Routing Replay (R3) for stable MoE training
  • Token-in-Token-Out (TITO) data path for multi-turn agentic sessions
  • Extensive hardware support including NVIDIA Blackwell/Hopper and AMD Instinct
  • Support for diverse RL algorithms including GRPO, GSPO, PPO, and REINFORCE++
  • Unified low-precision pipeline supporting FP8, MXFP8, and NVFP4

Related Releases

MiroThinker

MiroThinker is an open-source search agent model developed by MiroMindAI, designed for tool-augmented reasoning and real-world information seeking. It aims to match the deep research capabilities of leading AI models like OpenAI's Deep Research and Google's Gemini Deep Research.

MiroMindAI
Open

Letta Code SDK

The Letta Code SDK is a software development kit that enables developers to build deeply personalized agents with persistent memory that learn over time. It serves as the interface to Letta Code, facilitating the creation of stateful agents capable of continuous learning and improvement.

Letta AIJan 28
Open

OB-1

OB-1 is a self-improving coding agent developed by OpenBlock Labs, designed to autonomously handle the full development lifecycle, from project management to pull requests. It integrates seamlessly into existing workflows, enhancing productivity and code quality.

OpenBlock LabsSep 1
Open

AIO Sandbox

AIO Sandbox is an integrated environment designed for AI agents, combining a browser, terminal, filesystem, VSCode, Jupyter, and MCP Server into a single Docker container. This unified setup allows seamless development and execution of AI agents without the need for multiple services.

Agent InfraOct 29
Open

Discussion

0
Upvotes
0
Downvotes
0 reviews

Sign in to leave a review

Reviews

No reviews yet. Be the first to review!

🚀 Join the AI dev community — follow us everywhere

© 2026 MARKTECHPOST AI MEDIA INC. All rights reserved.Terms & ConditionsPrivacy Policy
Beta Mode