Question 1

How many AI models does OrcaRouter support?

Accepted Answer

OrcaRouter ships with 200+ models across OpenAI, Anthropic, Google, AWS Bedrock, Azure, Mistral, Meta Llama, DeepSeek, Qwen, Moonshot Kimi and ByteDance Seedance, all callable through one OpenAI-compatible API key.

Question 2

How quickly are new model releases added to the catalog?

Accepted Answer

Frontier models (GPT, Claude, Gemini) typically land in the OrcaRouter catalog within 24 hours of vendor release. Open-weights models (Llama, Mistral, Qwen, DeepSeek) usually within a few days, depending on upstream availability.

Question 3

Are model prices on this page live?

Accepted Answer

Yes. Per-token prices reflect the live upstream rate. OrcaRouter does not mark up provider pricing — what you see here is what your account is billed at request time.

Question 4

Can I filter models by modality (text / image / audio / video)?

Accepted Answer

Yes. Use the modality tabs above the catalog to scope to text, image generation, audio (TTS / STT), or video models.