AI Providers & Models

14 providers. All models auto-update to their latest version daily. Add your own API keys or use platform defaults.

Always on the Latest Model — No Config Needed

The platform polls every provider's API daily at midnight UTC. When a new model drops — GPT-5, Claude 5, Gemini 3, Grok 4 — your agents upgrade instantly. You always run the most advanced model available. Add your own API key for any provider to use your own quota and rate limits.

Hermes Agent

1 model · 1 agent

The brain and engine of Sahjony Capital. Autonomous orchestration, reasoning, and execution across all providers.

Autonomous orchestrationMulti-agent coordinationSelf-healing workflows+3
Platform default
View Models →
🟢

NVIDIA NIM

1 model · 1 agent

GPU-accelerated inference. Fastest execution for real-time trading signals. Free 1000 credits on signup.

GPU-accelerated inferenceSub-100ms latencyNVIDIA DGX Cloud+3
Platform default
View Models →
🔲

OpenAI

2 models · 2 agents

GPT series — the industry standard. o-series for deep reasoning, GPT-4o for speed.

Function callingStructured outputVision+2
Platform default
View Models →
🟤

Anthropic (Claude)

1 model · 1 agent

Claude — exceptional at analysis, research, and long-context reasoning. Safest for compliance.

200K contextExtended thinkingTool use+2
Platform default
View Models →
🔵

Google (Gemini)

1 model · 1 agent

Gemini — multimodal native. 1M+ context for full portfolio analysis in a single pass.

1M+ context windowNative multimodalGrounded search+2
Platform default
View Models →

xAI (Grok)

1 model · 1 agent

Grok — real-time knowledge from X. Unique edge for sentiment-driven trading.

Real-time X/Twitter dataFunction callingVision+2
Platform default
View Models →
🔵

Meta (Llama)

1 model · 1 agent

Llama — open-weight leader. Deploy on-premise for maximum data privacy.

Open weightsOn-premise deployFine-tunable+2
Platform default
View Models →
🟠

Mistral

1 model · 1 agent

Mistral — European excellence. Top-tier performance at competitive pricing.

Function callingJSON modeEmbeddings+3
Platform default
View Models →
🌀

DeepSeek

1 model · 1 agent

DeepSeek — frontier reasoning at 10x lower cost. Breakthrough price-performance.

Deep reasoning chainMoE architecture10x cost reduction+2
Platform default
View Models →
🟣

Cohere

1 model · 1 agent

Cohere — enterprise search and RAG. Best for knowledge-grounded analysis.

Enterprise RAGGrounded generationEmbeddings+3
Platform default
View Models →
🐙

OpenClaw

1 model · 1 agent

OpenClaw — open-source multi-agent framework. Build custom autonomous agents.

Multi-agent orchestrationTool-use chainsOpen-source+3
Platform default
View Models →

Groq

1 model · 1 agent

Groq — LPU inference. Sub-millisecond token speeds for ultra-low-latency trading.

Sub-ms token latencyLPU hardwareReal-time streaming+2
Platform default
View Models →
🔮

Perplexity

1 model · 1 agent

Perplexity — real-time web-grounded answers. Best for market research and news.

Real-time web searchCitationsMarket research+1
Platform default
View Models →
🦬

FreeBuff

2 models · 1 agent

FreeBuff — autonomous AI agent for free-tier model aggregation. Access premium models at zero cost via community-shared inference pools.

Free-tier model accessCommunity inference poolsMulti-model routing+4
Platform default
View Models →