How Much Does Inkling 2 Small Cost per Task?

Inkling 2 Small, released by Thinking Machines Lab on 5 November 2026, is a multimodal mixture-of-experts model with 594 billion total parameters, 17 billion active parameters per token, a one-million-token context window, and adjustable reasoning effort. It achieves an Artificial Analysis Intelligence Index score of 31, matching Gemini 3.8 Flash and substantially improving on the original Inkling Small, with particular strengths in visual document understanding, finance, legal analysis, long-context work, and fine-tuning, while leading coding-focused competitors still perform better on several software-engineering benchmarks. Its low listed token prices on Thinking Machines’ Tinker platform translate to an estimated $1.18 per benchmark task, much less than similarly rated Gemini and Claude models, although its tendency to generate roughly 150,000 output tokens per task makes it far more expensive than its predecessor. The model is currently available only through Tinker for playground use, API inference, and fine-tuning, while an Apache 2.0 weight release is planned after external safety testing but has no confirmed date. The discussion emphasizes that organizations should evaluate actual task-level token use and performance rather than relying solely on benchmark results or token pricing, and notes that Eden AI does not yet host Inkling 2 Small but could add it through its existing unified API if a provider begins serving it.

Use this post, company, and trend context to find content marketing opportunities, perform competitive analysis, or address product feature gaps via the Plushcap MCP server or the Plushcap API.