Nox 2: the next model generation
Product
Meet the Nox 2 generation: Orion 2, the new Orion Flash 2, and Vela 2, with 1.3 Astra unchanged.
Orion 2 upgrades the flagship
Nox Orion 2 is the upgrade to 1.4 Orion. It keeps everything 1.4 could do while becoming faster overall, more accurate, and lower priced.
The usage-equivalent reference price is $7 per 1M input tokens and $22 per 1M output tokens, compared with 1.4 Orion's $9.80 per 1M input tokens and $30.80 per 1M output tokens.
Orion Flash 2 is the new fast Orion
Nox Orion Flash 2 is new: the fast version of Orion. It is a 300-billion-parameter model that keeps Orion's full thinking depth while answering faster and at a lower price.
Its usage-equivalent reference price is $0.60 per 1M input tokens and $2.00 per 1M output tokens. Orion Flash 2 is available in Plan mode on Pro and MAX and in NOX WORK; it does not include Hyper Ultra.
Vela 2 for every plan
Nox Vela 2 is the everyday model on every plan. It is more accurate than every previous Vela while keeping the same speed.
Its usage-equivalent reference price is $0.15 per 1M input tokens and $0.60 per 1M output tokens.
Astra remains unchanged
Nox 1.3 Astra is unchanged. The Nox 2 generation updates Orion and Vela while keeping Astra available for deep reasoning.
Agent 1.2 and Light 1.1 are retired
Nox Agent 1.2 and its dedicated research mode are retired. Plan mode with live web access is now the way to get cited research.
Nox Light 1.1 is retired from consumer products, with Vela 2 as its successor. Light 1.1 remains available to developers through the Nox developer API.