◆ Models

Top AI models, one gateway

Mix and match flagship and fast models. Switch with a single string change, no SDK swap needed.

Tier
Provider

Claude Opus 4.7

Anthropic
flagship

Anthropic's top-tier model. Best for deep reasoning, agentic workflows, and long-form analysis.

In $15.00Out $75.00Speed 42 t/s
200k context·32k output

Claude Sonnet 4.6

Anthropic
flagship

Balanced flagship from Anthropic. Strong coding, writing, and tool use at workhorse pricing.

In $3.00Out $15.00Speed 78 t/s
200k context·16k output

Claude Haiku 4.5

Anthropic
fast

Fast and inexpensive. Ideal for chatbots, classification, and structured extraction at scale.

In $0.80Out $4.00Speed 156 t/s
200k context·8k output

Minimax 2.7

Minimax
flagship

Latest Minimax flagship with strong multilingual and multimodal capabilities.

In $1.20Out $4.80Speed 92 t/s
256k context·16k output

Minimax 2.5

Minimax
standard

Reliable, cost-effective Minimax model for general chat and content generation.

In $0.60Out $2.40Speed 124 t/s
256k context·8k output

Deepseek V4 Pro

Deepseek
flagship

Top Deepseek model. Excels at reasoning, math, and repo-scale code generation.

In $0.55Out $2.19Speed 88 t/s
128k context·16k output

Deepseek V4 Flash

Deepseek
fast

Lightweight Deepseek variant. Fast, cheap, and capable for high-volume tasks.

In $0.14Out $0.28Speed 198 t/s
128k context·8k output

GLM 5.1

Zhipu
flagship

Zhipu's flagship general-purpose model with strong Chinese and English performance.

In $0.50Out $2.00Speed 95 t/s
128k context·8k output

GLM 5

Zhipu
standard

Solid general-purpose GLM model. Reliable for chat, summarization, and tool calling.

In $0.30Out $1.20Speed 132 t/s
128k context·8k output

GPT 5.3 Codex

OpenAI
flagship

OpenAI's specialized coding model. Built for refactoring, debugging, and agentic dev workflows.

In $5.00Out $20.00Speed 64 t/s
256k context·16k output

GPT 5.5

OpenAI
flagship

OpenAI's most advanced general-purpose flagship. Multimodal, strong reasoning, broad knowledge.

In $8.00Out $32.00Speed 56 t/s
400k context·32k output

GPT 5.4

OpenAI
flagship

Previous OpenAI flagship. Excellent reasoning and tool use, slightly cheaper than 5.5.

In $4.50Out $18.00Speed 72 t/s
256k context·16k output