Nex-N2.6

Nex-N2.6 is an agentic model from Nex AGI. It showed up in this scan as a new OpenRouter sitemap URL, and the model page lists the release date as 10 November 2026. The model id is nex-agi/nex-n2.6. OpenRouter describes it as built on GLM-5.3-Flash, with about 321 billion parameters, and aimed at computer use, coding, and long-horizon tasks that involve tools.

Compared with Nex-N2.5-Pro, the page says it reduces repetitive reasoning and follows the tools that are actually available in the current environment more closely. The context window is 262,144 tokens. Inputs are text and images; output is text. Tool calling and structured outputs are listed as supported. This page did not find a separate same-day Nex AGI blog post, so the OpenRouter model page is the primary source used here.

Highlights:

• Model page: https://openrouter.ai/nex-agi/nex-n2.6 ;
• Release date on that page: 10 November 2026;
• Listed price: $0.12 per million input tokens and $0.40 per million output tokens, plus a cache-read rate; the amount you actually pay can be lower with discounts and cache hits;
• The same lab also lists Nex-N2.5-Mini and Nex-N2.5-Pro; the Nex family was not already in this radar database.

Why it matters

Recent agent-model updates are less about whether a model can call a tool and more about whether it loops on unavailable tools or repeats the same chain of thought. Nex-N2.6 is positioned there: a GLM-5.3-Flash-based model for computer use and long-horizon tool work, with two explicit differences versus N2.5-Pro written on the listing. If you already route through OpenRouter or a Nex AGI endpoint, this is a new SKU name that appeared today, not an unverified rumor.

The same page publishes a 262K context window and a public list price, which is enough to sketch a cost comparison against other coding or computer-use models. In this radar window the signal was a newly added OpenRouter sitemap page dated today.

Limits

The public description is mainly the OpenRouter model page. This write-up did not confirm a matching long-form Nex AGI announcement, so parameter count, less repetitive reasoning, and task fit are vendor or platform claims, not independent benchmarks. Prices move with discounts, cache hits, and routing; the page also shows a weighted average of amounts actually paid. Because the model is built on GLM-5.3-Flash, license, data, and safety terms come from both Nex AGI and the base model. Do not treat unpublished scores or blanket superiority claims as facts.

1. Open https://openrouter.ai/nex-agi/nex-n2.6 and confirm the current price, context length, and provider status;
2. Set the model id to nex-agi/nex-n2.6 on OpenRouter (or another gateway that already offers Nex AGI) and run a real computer-use or in-repo coding task with an actual tool list;
3. Compare against Nex-N2.5-Pro on whether it repeats itself less and whether it only calls tools that exist in the current environment;
4. Before production use, check rate limits, cache pricing, and the license plus data-handling terms for both Nex AGI and the GLM base model.