MoonshotAI: Kimi K2.5

Kimi K2.5 is a multimodal model from Moonshotai that accepts text and image inputs, supports tool use and reasoning, and offers a 262,144-token context window. Structured output support is unconfirmed. The context length puts it in a range suitable for long-document work, codebases, or extended conversations, and tool and reasoning support makes it viable for agentic workflows. On pricing, at $0.375 per million input tokens and $2.025 per million output tokens it sits in a mid-range tier, neither the cheapest option nor a premium outlier. Its blended benchmark score of 73.3 is drawn from only two benchmarks, so treat that figure as preliminary rather than authoritative. Teams doing long-context reasoning or image-plus-text tasks who want a competitively priced model should shortlist it, but those needing well-validated performance data may want to wait for broader benchmark coverage before committing.