What is the A2A (Agent2Agent) protocol?

A2A is the open protocol that lets agents built on different frameworks call each other. Agents publish an Agent Card describing their skills, and other agents read that card to decide whether and how to invoke them.

How do I make my existing agent A2A compliant?

Run super a2a adapt --entrypoint : --framework . SuperOptiX reads your agent's structure, derives its skills, and writes an A2A Agent Card plus a conformant server. Your agent code is not modified.

Which agent frameworks does SuperOptiX support for A2A?

Eight: DSPy, CrewAI, the OpenAI Agents SDK, Pydantic AI, Google ADK, the Claude Agent SDK, DeepAgents and the Microsoft Agent Framework. Adaptation works on agents written directly against those libraries, whether or not SuperOptiX built them.

How conformant is the SuperOptiX A2A implementation?

It records zero failures against the official A2A Technology Compatibility Kit. Every requirement the TCK exercises against the conformance harness passes: 73 of 73 at MUST, 7 of 7 at SHOULD, 4 of 4 at MAY. The TCK also prints a headline percentage of 77.7% at MUST, which counts 25 requirements it cannot exercise here, covering authentication, TLS, Agent Card signing and gRPC.

How do I make my agent more discoverable to other agents?

Optimize the Agent Card text. SuperOptiX scores discovery rate, invocation rate and confusion rate over a query set, then uses GEPA to rewrite skill names, descriptions and tags against that score. Agent behaviour is unchanged.

Is SuperOptiX free and open source?

Yes, under the MIT License. The A2A implementation, the adapt command, the conformance suite, the routing optimizer, the SuperSpec compiler and all eight framework integrations are in the open source package. Install with uv tool install superoptix.

Is there a live A2A agent I can call?

Yes. A SuperOptiX agent runs at https://a2a.superoptix.ai on Google Cloud Run, with its Agent Card published at https://superoptix.ai/.well-known/agent-card.json.

Who builds SuperOptiX?

Superagentic AI, which builds agentic AI infrastructure and offers support and custom engineering engagements.

Unified Model Management

One CLI for All Models

Local models, cloud APIs, production inference. SuperOptiX unifies them all. Install, run, and manage models across Ollama, MLX, HuggingFace, vLLM, SGLang with a single interface.

Local Backends

100+

Cloud Providers

Production Engines

Unified CLI

Local Inference

Local Model Backends

Run models locally for privacy, speed, and cost savings. Choose the backend that fits your platform.

Recommended

Ollama

Easiest for Everyone

Ease of Use

The simplest way to run local models on any platform. Perfect for beginners with automatic server management.

Platforms

Windows

macOS

Linux

Default Port

11434

One-command installation

Auto-start server

GPU acceleration

Model library built-in

Best For:

Beginners, All platforms

MLX

Apple Silicon Native

Ease of Use

Apple's native ML framework for blazing-fast inference on M1/M2/M3 Macs. Up to 5x faster than alternatives.

Platforms

Apple Silicon Only

Default Port

8000

Native Apple Silicon

Ultra-fast inference

Memory efficient

GPT-OSS support

Best For:

Apple Silicon users

LM Studio

Visual Interface

Ease of Use

User-friendly desktop GUI for model management. Browse, download, and run models with a beautiful interface.

Platforms

Windows

macOS

Default Port

1234

Visual model browser

Easy configuration

Chat interface

GGUF support

Best For:

Windows users, GUI lovers

HuggingFace

Maximum Flexibility

Ease of Use

Access thousands of open-source models directly. Best for advanced users who need custom configurations.

Platforms

Windows

macOS

Linux

Default Port

8001

Largest model library

Custom models

Research models

Fine-tuning ready

Best For:

Advanced users, Researchers

Cloud APIs

Cloud Model Providers

Access 100+ cloud providers through LiteLLM integration. From OpenAI to open-source alternatives.

OpenAI

Models

GPT-4o

GPT-4o-mini

GPT-3.5-Turbo

Best For

General purpose, coding

OPENAI_API_KEY

Anthropic

Models

Claude-3-Opus

Claude-3-Sonnet

Claude-3-Haiku

Best For

Reasoning, analysis

ANTHROPIC_API_KEY

Google

Models

Gemini-Pro

Gemini-Pro-Vision

Best For

Multimodal capabilities

GOOGLE_API_KEY

Groq

Models

Llama-3-70B

Mixtral-8x7B

Best For

Ultra-fast inference

GROQ_API_KEY

Together AI

Models

Llama-3.1-405B

Mixtral

Open models

Best For

Open source models

TOGETHER_API_KEY

Mistral

Models

Mistral-Large

Mistral-Small

Best For

European AI models

MISTRAL_API_KEY

Cost Warning: Monitor cloud usage carefully during optimization. Costs can add up quickly!

Production Ready

Production Inference Engines

Scale to production with high-performance inference servers. Handle thousands of concurrent requests.

vLLM

High-throughput serving with PagedAttention for production deployments

PagedAttention

Continuous batching

Multi-GPU

Quantization

Requirements

Linux + NVIDIA GPU

SGLang

Streaming-optimized inference with RadixAttention for real-time applications

RadixAttention

Streaming generation

High concurrency

Prefix caching

Requirements

Linux + NVIDIA GPU

TGI

HuggingFace's Text Generation Inference for easy model serving

Easy deployment

Tensor parallelism

Watermarking

Token streaming

Requirements

Linux + NVIDIA GPU

Learn about vLLM & SGLang setup

Super CLI

One CLI for Everything

Discover, install, run, and manage models across all backends with intuitive commands.

$ super model listList all installed models

$ super model list --allList all available models

$ super model discoverGet model discovery guide

$ super model list --backend ollamaFilter by backend

Quick Start: Run Your First Model

# Install SuperOptiX

$ uv pip install superoptix

# Install a model (auto-detects best backend)

$ super model install llama3.2:3b

# Run the model

$ super model run llama3.2:3b "Write a Python hello world"

🎉 Model running with auto-installation if needed!

Coming Soon

Model Intelligence

The next evolution of model management. AI-powered recommendations, performance analytics, and automatic optimization. Coming to SuperAgents tier.

Smart Model Selection

AI-powered recommendations based on your use case, hardware, and requirements

Performance Analytics

Detailed metrics on tokens/sec, memory usage, response time, and quality

Cost Optimization

Automatic budget monitoring and recommendations to reduce cloud costs

Resource Management

Monitor GPU/CPU usage, memory consumption, and optimize allocation

🔮 Preview: Model Intelligence Commands

super model discover --use-case code

super model quantize llama3.2:3b

super model info llama3.2:3b

super model list --all

📅 Expected: Later 2025 • SuperAgents Tier

Why SuperOptiX for Model Management

❌ Other Frameworks

No direct model execution - require server setup

Manual model installation and configuration

No unified interface across backends

Limited cloud provider integration

No automatic model discovery

Complex server management

SuperOptiX

Direct model execution with auto-installation

One-command model installation and setup

Unified CLI for all backends (Ollama, MLX, HF, etc.)

Seamless cloud provider integration (100+ via LiteLLM)

Automatic model discovery and recommendations

Built-in server management and auto-configuration

Ready to Manage Models Effortlessly?

Start with local models for free, scale to cloud when needed. One CLI, all models.

🔧 Model Management •☁️ Cloud Inference •🧠 Model Intelligence