Which models are available and how do I reference them?
Which models are available and how do I reference them?
provider/ prefix, for example openai/gpt-5.6-sol, anthropic/claude-sonnet-5, or google/gemini-3.5-flash. Models must be enabled on the Models page before they can be called.Do I need my own provider API keys?
Do I need my own provider API keys?
Can I use private or custom models?
Can I use private or custom models?
How do I find a model's context window or max output tokens?
How do I find a model's context window or max output tokens?
Why do most chat models require a user message?
Why do most chat models require a user message?
400 invalid_input error. Tool results are grouped into user turns automatically.How should I configure temperature and top_p?
How should I configure temperature and top_p?
0 to 1, OpenAI models 0 to 2. Newer Anthropic models reject requests that set temperature and top_p together; set one or the other. Audio endpoints accept 0 to 1. See the Anthropic provider page and Reasoning.How do reasoning models behave?
How do reasoning models behave?
reasoning_effort, while Google Gemini and Anthropic use a thinking object. When reasoning_effort is set, the AI Gateway drops temperature and top_p before forwarding the request, since those parameters are incompatible with OpenAI reasoning models. See Reasoning.What happens when I send unsupported or out-of-range parameters?
What happens when I send unsupported or out-of-range parameters?
4028 when temperature and top_p are both specified. See Reasoning for the parameters the gateway drops on reasoning models.When does native file attachment work?
When does native file attachment work?
Which models support image, PDF, or audio input?
Which models support image, PDF, or audio input?
Which models support structured output?
Which models support structured output?
Which models support streaming?
Which models support streaming?
Which models support tool calling?
Which models support tool calling?
Which models support prompt caching?
Which models support prompt caching?
cache_control markers, Google Gemini models cache implicitly, and OpenAI models cache automatically. See Prompt Caching.Which model should I pick, or can Orq.ai choose per request?
Which model should I pick, or can Orq.ai choose per request?
If a provider fails or rate-limits, can traffic fail over?
If a provider fails or rate-limits, can traffic fail over?