Models

0G.AI in-house model optimized for agentic coding and tool use; thinking enabled by default.

Zhipu AI flagship purpose-built for long-horizon tasks; 744B foundation, 200K context.

Multilingual automatic speech recognition (ASR); transcription and translation.

Asynchronous text-to-image model with Base64 output. Generates at most 2 images per request — requesting more (n > 2) returns 2 images, not an error.

671B mixture-of-experts model with hybrid thinking; strong at coding, math, and multi-step reasoning.

Lightweight MoE (284B total / 13B active) with native 1M context; low-latency, low-cost.

DeepSeek flagship for agentic coding, multi-step workflows, and complex reasoning; 1M context, up to 384K output.

Next-generation model purpose-built for coding and agent workflows; 744B foundation.

Frontier natively-multimodal model on MiniMax Sparse Attention (MSA); agentic coding, native tool use, and long-horizon tasks. 1M context, thinking on by default.

Alibaba multimodal vision-language model; strong at visual reasoning, OCR, and document understanding.

Alibaba flagship with hybrid linear attention and sparse MoE; 1M context, 119 languages.

Alibaba flagship with native function calling and web search; 1M context.

Alibaba multimodal model with vision and video understanding; native function calling, 1M context.