As of August 18, 2026, the unified board aggregates public model listings (e.g., OpenRouter) to compare context, modality, and average price per million tokens. Leaders in this snapshot include "SpaceXAI: Grok 4.20 Multi-Agent", "DeepSeek V4 Flash Latest", "Meta: Llama 4 Scout", "DeepSeek: DeepSeek V4 Flash 0731", "Xiaomi: MiMo-V2.5", "OpenAI: GPT-5.6 Luna Pro (batch)". Across the top 12, kinds skew toward Multimodal, LLM, Agent, reflecting sustained demand for long-context and multimodal APIs.
Versus the prior snapshot, about 1 row(s) show upward price/metric movement and 3 downward (see arrows in-table). Among these rows, average USD/1M tokens spans roughly $0.00–$8.75 before provider-specific discounts.
For teams choosing APIs, rank is only one signal—latency, safety policies, and regional availability matter as much as list price.