Amazon: Nova Lite 1.0

Amazon Nova Lite 1.0 is a multimodal model from Amazon that accepts both text and image inputs. It offers a 300,000-token context window, supports tool use, and caps output at 5,120 tokens per response. It does not support reasoning modes, and structured output support is unconfirmed, so buyers who need those features should look elsewhere before committing. At $0.06 per million input tokens and $0.24 per million output tokens, it sits at the budget end of multimodal pricing, which makes it worth considering for high-volume document or image processing workloads where cost control matters. The caveat is real: there is no independent benchmark coverage to validate quality claims, so teams with quality-sensitive use cases are essentially evaluating it blind. Developers running low-stakes, cost-sensitive pipelines may find it worth trialing, but those who need proven benchmark performance should wait for third-party results.