50+ frontier models. One Token. Switch on the fly.

GPT-5

OpenAI

Flagship

OpenAI's flagship reasoning + agentic model. Best-in-class for long planning, code refactor and tool use.

GPT-4o

OpenAI

Fast

Multimodal text + vision + audio at half the cost of flagship — ideal for high-volume streaming chat.

Claude 4.5 Opus

Anthropic

Flagship

Anthropic's deepest thinker. Million-token context, ideal for long reasoning and multi-step agent workflows.

Claude 4.5 Sonnet

Anthropic

Fast

Fastest Claude model that still codes like Opus. The default pick for high-throughput coding agents.

Gemini 2.5 Pro

Google

Flagship

Google's 1M-context multimodal flagship. Best at parsing huge codebases and long PDFs.

Gemini 2.5 Flash

Google

Fast

Sub-second latency at <1¢/M tokens — perfect for Agent loops and bulk classification.

Grok 4

xAI

Flagship

xAI's latest with first-class web search and real-time data access for current events.

Meta's open-weight 405B mixture-of-experts. Run it via TonBo or self-host with the same API.

Open-weight reasoning model competitive with GPT-4-class for code generation at a fraction of the cost.

Alibaba's flagship — strong on Chinese workloads and competitive with closed models.

European-built flagship with strong code & math, multilingual native.

Cohere's RAG-optimized model with native tool use and citation grounding.

Midjourney V7

Midjourney

Vision

Industry-leading image generation routed through TonBo's stable tunnel.

Stable Diffusion 4

Stability AI

Vision

Open-weight image generation, perfect for batch generation pipelines.

Perplexity Sonar

Perplexity

Fast

Live web search with sources cited, ideal for fact-grounded chatbots.

GitHub Copilot

GitHub

Fast

GitHub's code-completion model — TonBo proxies it stably across every region.