All Frontier Models in One Place

Different tasks need different models. Use all of the leading frontier models in one place on Aurora.

Models on Aurora

The list of models and providers available on Aurora is shown below

Auto

Auto is Aurora’s automatic routing. It picks the right frontier model for each step of a build, favoring faster models for routine edits and stronger reasoning when the task needs it, so you don’t have to choose a lab yourself.

Available on Aurora

  • Auto

    auto

Anthropic

Anthropic builds Claude, a family of AI systems focused on helpful, honest, and harmless assistance. The lab is known for careful alignment research, strong coding and reasoning performance, and models that excel at following complex instructions across long contexts.

Anthropic model docs

Available on Aurora

Loading models…

OpenAI

OpenAI develops the GPT family and related systems used widely for coding, analysis, and agent workflows. Their research spans reasoning, multimodality, and tooling — with a broad ecosystem of APIs and developer documentation for choosing the right capability tier.

OpenAI model docs

Available on Aurora

Loading models…

Gemini

Gemini is Google DeepMind’s frontier model family, built for multimodal understanding and very large context windows. The line emphasizes speed, long-document reasoning, and tight integration with Google’s AI research and developer platform.

Gemini model docs

Available on Aurora

Loading models…

SpaceXAI

SpaceXAI builds the Grok model family with an emphasis on real-time knowledge, high-throughput inference, and capable coding assistants. Their docs compare Grok variants so you can balance latency, depth, and cost.

xAI model docs

Available on Aurora

Loading models…

Mistral

Mistral AI is a European frontier lab known for efficient, high-performance open-weight and API models. They specialize in strong multilingual support, practical coding assistants, and a clear tiering of model sizes for different latency and quality targets.

Mistral model docs

Available on Aurora

Loading models…

DeepSeek

DeepSeek builds competitive frontier models with a focus on reasoning and coding efficiency. Their systems are widely used for cost-effective agent workloads, with documentation that outlines how their chat and reasoner-style offerings differ.

DeepSeek docs

Available on Aurora

Loading models…

Meta

Meta develops the Llama family of open and hosted models, spanning multimodal understanding and efficient inference. Llama models are widely used for coding assistants, research, and production agents where openness and flexibility matter.

Meta model docs

Available on Aurora

Loading models…

Moonshot AI

Moonshot AI builds the Kimi model family, known for very long context windows, strong coding and agent performance, and multimodal understanding. Their platform documents how Kimi variants balance depth, speed, and tool use.

Kimi model docs

Available on Aurora

Loading models…

ZAI

ZAI develops the GLM model family with a focus on coding, agent workflows, and long-horizon reasoning. Their docs cover flagship GLM releases and how to choose between capability tiers for software engineering tasks.

ZAI model docs

Available on Aurora

Loading models…

Model access scales with your plan

Plus unlocks all frontier models. See plans for model accessibility options.