Nemotron Cascade 2 30B A3B
Analysis Summary
NVIDIA's Nemotron Cascade 2 30B A3B has a mixed benchmark profile. Its instruction-following result is strong, and its terminal-task measure is the most useful technical signal in this set of models. The coding and agentic indices remain low, while its general intelligence result suggests that performance may vary substantially across task types. Pricing and context-window information are not supplied.
The results make it a candidate for testing on structured prompts and bounded terminal-oriented tasks, but not a proven choice for autonomous coding agents or broad client analysis. Strong instruction following can help with repeatable workflows, yet reliability should be checked on the actual prompts and tools the business intends to use. With no pricing data, cost-effectiveness is unknown. Keep it in a task-specific evaluation rather than putting it on a client stack by default.
Assessed September 23, 2026
Editorial notes
Nemotron Cascade 2 30B A3B combines strong instruction-following and terminal-task results with a low overall intelligence measure; coding and agentic indices remain limited, and pricing is not provided.
The DFO score ranks models on the Artificial Analysis Intelligence Index and is refreshed as new models launch. The write-up and verdict are our own. Feedback?
DFO Verdict
Nemotron Cascade 2 30B A3B combines strong instruction-following and terminal-task results with a low overall intelligence measure; coding and agentic indices remain limited, and pricing is not provided.
Benchmarks
Artificial Analysis data refreshed Oct 15, 2026. Arena ratings from the Oct 13, 2026 text leaderboard, with style control. Bars: magenta = reasoning, ink = coding and agents, cyan = instructions and long context.
How Nemotron Cascade 2 30B A3B compares
Nemotron Cascade 2 30B A3B ranks #205 of 439 AI models we track for overall intelligence, #120 of 198 for coding, #121 of 181 for agentic tasks. Nemotron Cascade 2 30B A3B is currently free to use via OpenRouter.
Dark bar = input · light bar = output, scaled to the priciest peer.
1M tokens ≈ 8 full-length novels or ~2,500 pages of business documents in a single request.
Frequently asked questions about Nemotron Cascade 2 30B A3B
How much does Nemotron Cascade 2 30B A3B cost?
Nemotron Cascade 2 30B A3B is currently available for free via OpenRouter.
Is Nemotron Cascade 2 30B A3B good for coding?
On our coding benchmark index, Nemotron Cascade 2 30B A3B ranks #120 of 198 models, placing it in the broader range of the field for code generation and debugging.
Who created Nemotron Cascade 2 30B A3B?
Nemotron Cascade 2 30B A3B is developed by NVIDIA and was released on March 19, 2026.