Artificial Analysis lists Nova 2.0 Omni release
Artificial Analysis has added a Nova 2.0 Omni release page covering three models with a 1-million-token context window, though the page excerpt includes no benchmark scores, prices or speed figures.
Benchmarking firm Artificial Analysis has published a release page for Nova 2.0 Omni. The page says the release includes three models, each with "different intelligence, performance, and pricing characteristics."
The listed specifications are:
- Input: text and image
- Output: text only
- Context window: 1M tokens, which Artificial Analysis puts at about 1,500 A4 pages
Although the release is called "Omni," the page lists only text and image inputs. It does not mention audio or video.
The page has sections for intelligence versus cost per task, cost per Intelligence Index task, time per task, and capability indexes. The capability indexes draw on evaluations including AA-Briefcase v1.1, AA-Omniscience v1.1, HLE-Diamond, AutomationBench-AA, GDP.pdf, MLCR-AA, CritPt v2 (Hard Subset), Terminal-Bench 4.0, Terminal-Bench-Science 0.1 and FrontierCode-AA 2.0. The page compares models within their class. Reasoning models are measured against both reasoning and non-reasoning models. Proprietary models are compared with others in the same price band, using a blended 3:1 input/output token price.
The material provided does not include the actual scores, prices, speed results, names of the three models, or whether any of them are reasoning models. It also does not name the developer or give a release date.