Moonshot's 1T-param open-source agentic model
Category: Text & Copy. Pricing: current comparable pricing is not verified; check the vendor's website. Last modified: 2026-09-07.
Moonshot AI's flagship, a 1-trillion-parameter Mixture-of-Experts model with 32B active params. Open-source under modified MIT, with long-horizon autonomous execution and Agent Swarm (up to 300 sub-agents). Leads HLE-Full with tools at 54.0.
Not editorially tested: One AI Guide has not published a complete hands-on test record for this tool.
Published by LUMIVEX for One AI Guide.
Editorial lead: Said El Moussaoui. Our methodology is published at /how-we-rate.
Official vendor website: see the vendor link below.
These are sourced comparison facts, not hands-on test results. Partial facts need more confirmation; stale facts need a fresh check.
Yes
Kimi K2.6 Model PricingModel: kimi-k2.6; CheckedDate: 2026-08-27; ContextUnit: tokens; ContextWindow: 262144
Kimi K2.6 Model PricingNot currently confirmed
Kimi K2.6 is a text & copy tool. Moonshot's 1T-param open-source agentic model.
Kimi K2.6 is current comparable pricing is not verified; check the vendor's website. Always check the vendor's site for the latest plans before signing up.
The catalog currently records Kimi K2.6's pricing model as Freemium. Check the vendor's site for the current free access, limits, and paid-plan terms.
Common use cases include Agentic coding, Long-horizon tasks, and Research.
The One AI Guide catalog currently highlights these strengths: Open-source MoE 1T; Agent Swarm (300 sub-agents); Strong on SWE-Bench.