Sulat.com
AI models
$10 off the fastest DeepSeek V4.1 Flash, Kimi K3.1 Pro and GLM 5.3-Flash from Synthetic
OpenRouter logo

Model details

Nex-N2.6

Nex-N2.6 is Nex AGI’s agentic model, built by post-training GLM-5.3-Flash and carrying about 321 billion parameters. It is intended for computer use, coding, and long-horizon tasks that depend on tools. The model listing says that, relative to Nex-N2.5-Pro, it reduces repetitive reasoning and follows the tools available in the current environment more closely. That makes the release especially relevant to agents whose value depends on selecting the right tool, acting in context, and avoiding repeated reasoning loops rather than merely producing a polished answer.

In practice, Nex-N2.6 is best viewed as a tool-using model for workflows that can be checked against an environment, such as coding tasks, computer interaction, and extended agent runs. Its architecture change makes it more than a routine family update: moving to GLM-5.3-Flash changes the underlying tokenizer, chat template, tool-call interface, and likely failure patterns, so prompts and integration logic tuned for Nex-N2.5-Pro may need validation. Recruitly’s recruitment-work benchmark includes a Nex n2.6 entry with a quality score of 79, no hard failures, and 14.5-second latency, providing limited real-work evidence beyond the creator’s positioning claim. The most practical path is therefore a side-by-side evaluation on measurable agent tasks, watching tool adherence, repetition, task completion, and token use per completed workflow.

OpenRouternex-agi/nex-n2.6agi

Quick Info

Powered by
Provider
OpenRouter
Model key
nex-agi/nex-n2.6
Release date
Nov 10, 2026
Last updated
Nov 10, 2026
Input modalities
Output modalities
Capabilities

Cost

Input token cost
$0.12
Output token cost
$0.40

Limits

Output tokens
235,929 tokens
Context window
262,144 tokens

Transparent token rates

Compare Nex-N2.6 pricing

Rates are shown per one million tokens. Combined means one million input plus one million output tokens.

Browse this family

Latest news about Nex-N2.6

OpenRouter

CoverageRelease Notes

Price Per Token’s November 10, 2026 release tracker lists Nex AGI’s Nex-N2.6 under the exact identifier nex-agi/nex-n2.6, published through OpenRouter. The listing credits Nex AGI as the model creator and OpenRouter as the distribution listing. The entry corroborates the model’s recent appearance, but the supplied page provides no capabilities, architecture, context length, benchmarks, API changes, or pricing details. It is a third-party release tracker rather than a Nex AGI announcement or model card.

OpenRouter

CoverageBenchmark

Recruitly's benchmark of 49 models on real recruitment workloads ranks Nex AGI's Nex n2.6, accessed via OpenRouter, at ninth place with a quality score of 79 and zero hard failures. That places it just below Moonshot AI's Kimi K3.1 Pro at 80 and level with Google's Gemini 4 Argon at 79. The listed figures show a 262,144-token context window, $0.12 input and $0.40 output pricing, and 14.5-second latency, with a very low $0.0004 cost per task. The comparatively low latency and per-task cost give the model practical appeal for agentic automation workloads that need a long context window.

OpenRouter

CoverageBenchmark

Nex-N2.6 went live on Nex AGI's own commercial API on November 10, 2026 with no launch post, no readable model card, and no benchmark table, according to this analysis. The post-training base changed from Nex-N2.5-Pro's Qwen3.5 mixture-of-experts checkpoint to GLM-5.3-Flash, Z.ai's 320-billion-parameter multimodal model, yielding roughly 321B parameters. The model also launched at a higher price than the version it supersedes, at $0.12 per million input tokens, a notable step up for the Nex line. The write-up frames the release as defined by a base-model switch and price increase rather than by published evaluation results, which limits how well the agentic claims can be independently checked.

OpenRouter

Coverage

OpenRouter lists Nex-N2.6 as an agentic model from Nex AGI, built on a GLM-5.3-Flash base with roughly 321B parameters and a 262K context window, released November 10, 2026. It targets computer use, coding, and long-horizon tool workflows, and versus Nex-N2.5-Pro it is described as reducing repetitive reasoning and adhering more closely to the tools present in the environment. Pricing on the model page is $0.12 per million input tokens and $0.40 per million output tokens, shown with a 20% promotion, plus a $0.03 per million cache-read rate. Reported performance includes about 90 tokens per second throughput with a 0.80-second P50 latency, 100% uptime, and a low 0.19% tool-call error rate for the single listed host, Nex AGI.

Sulat also reports on AI and tech beyond the catalog.Read news.sulat.com

Videos about Nex-N2.6

More models around Nex-N2.6