ReasonKit Integrations

OpenAI

GPT-5.2, o3, GPT-4-Turbo

Full support for OpenAI's function calling API. Works with all GPT models including the latest reasoning models.

Function Calling JSON Mode Streaming

# Quick start

              export OPENAI_API_KEY="sk-..."

              reasonkit analyze
              "Your question" --provider
              openai

Anthropic

Claude Opus 4.5, Sonnet 4.5, Haiku

Native tool use support with Claude models. Extended context windows perfect for deep analysis and multi-document reasoning.

Tool Use 200K Context Vision

# Quick start

              export ANTHROPIC_API_KEY="sk-ant-..."

              reasonkit analyze
              "Your question" --provider
              anthropic

Google Gemini

Gemini 3 Pro, 3 Flash, 2.5 Pro

Leverage Google's multimodal reasoning with 1M+ token context. Excellent for document analysis and visual reasoning tasks.

1M+ Context Multimodal Grounding

# Quick start

              export GOOGLE_API_KEY="AIza..."

              reasonkit analyze
              "Your question" --provider
              gemini

Ollama

Llama 3.3, Mistral, Qwen, DeepSeek

Run structured reasoning entirely locally. No API keys, no data leaving your machine. Perfect for sensitive analysis.

100% Local No API Key Air-Gapped

# Quick start (no API key needed)

              ollama pull llama3.3

              reasonkit analyze
              "Your question" --provider
              ollama

Azure OpenAI

Enterprise GPT with compliance

Enterprise-grade deployment with Azure's compliance certifications. SOC 2, HIPAA, and GDPR compliant infrastructure.

Enterprise SOC 2 HIPAA

# Quick start

              export AZURE_OPENAI_ENDPOINT="..."

              export AZURE_OPENAI_KEY="..."

              reasonkit analyze "..." --provider
              azure

AWS Bedrock

Claude, Llama, Titan on AWS

Access multiple foundation models through AWS infrastructure. Native IAM integration and VPC support.

Multi-Model IAM Auth VPC

# Uses AWS credentials

              export AWS_REGION="us-east-1"

              reasonkit analyze "..." --provider
              bedrock

Groq

Fastest inference, LPU hardware

Lightning-fast inference on custom LPU hardware. Sub-100ms responses for rapid iteration and real-time analysis.

Ultra-Fast <100ms LPU

# Quick start

              export GROQ_API_KEY="gsk_..."

              reasonkit analyze "..." --provider
              groq

Together AI

Open models at scale

Access 100+ open-source models with OpenAI-compatible API. Cost-effective for high-volume structured reasoning.

100+ Models OpenAI Compatible Cost-Effective

# Quick start

              export TOGETHER_API_KEY="..."

              reasonkit analyze "..." --provider
              together

OpenRouter

340+ models, one API

Universal gateway to 340+ models from every major provider. Single API key for maximum flexibility.

340+ Models All Providers Unified API

# Quick start

              export OPENROUTER_API_KEY="sk-or-..."

              reasonkit analyze "..." --provider
              openrouter

Same Reasoning, Any Provider

ReasonKit's structured reasoning protocols work identically across all providers. Switch models without changing your workflow.

View Integration Docs