Supported Models

CallStack AI Gateway provides unified access to leading foundation models through a single, standardized API interface.

Anthropic Claude

> Gateway Endpoint: claude.callaiapi.com

🟠

Claude Opus 4

Most capable model for complex reasoning, code generation, and deep analysis tasks.

New
🟠

Claude Sonnet 4

Optimal balance of performance and cost for production workloads.

Production
🟠

Claude Haiku 3.5

High-speed, cost-efficient model for high-throughput processing tasks.

Production

OpenAI GPT Series

> Gateway Endpoint: openai.callaiapi.com

🟢

GPT-4o

Flagship multimodal model supporting text, image, and audio inputs.

Production
🟢

GPT-4o Mini

Lightweight, cost-optimized model for high-volume conversational workloads.

Production
🟢

o3 / o3-mini

Advanced reasoning models with built-in chain-of-thought for complex logic tasks.

New

Google Gemini

> Gateway Endpoint: gemini.callaiapi.com

🔵

Gemini 2.5 Pro

1M token context window with exceptional multimodal capabilities.

New
🔵

Gemini 2.0 Flash

Ultra-fast inference with low latency, ideal for real-time applications.

Production

Additional Models

> DeepSeek: deepseek.callaiapi.com · Meta: meta.callaiapi.com · Mistral: mistral.callaiapi.com

🟣

DeepSeek V3 / R1

High-performance open-source models with strong multilingual capabilities.

Production
🩷

Llama 3.1 405B

Meta's largest open model for enterprise-scale deployments.

Production
🔴

Mistral Large

European-built model with strong reasoning and multilingual performance.

Production

Model Capabilities Matrix

ModelEndpointContextMultimodalStreamingFunction Calling
Claude Opus 4claude.callaiapi.com200K
Claude Sonnet 4claude.callaiapi.com200K
GPT-4oopenai.callaiapi.com128K
o3openai.callaiapi.com200K
Gemini 2.5 Progemini.callaiapi.com1M
DeepSeek V3deepseek.callaiapi.com128K

::: tip Model Updates We continuously evaluate and onboard new models as they become available. For specific model requirements for your organization, contact our solutions team. :::