Nemotron 3 Ultra vs GLM 4.7
GLM 4.7 is cheaper on output tokens, while Nemotron 3 Ultra offers a larger context window. Choose Nemotron 3 Ultra or GLM 4.7 based on the trade-off between cost, context, and the benchmarks that matter for your use case.
| Spec | Nemotron 3 Ultra | GLM 4.7 |
|---|---|---|
| Provider | NVIDIA | Z.ai |
| Input / 1M tokens | $0.60 | $0.60 |
| Output / 1M tokens | $2.40 | $2.20 |
| Context window | 262K | 205K |
| Parameters | 550B | 358B |
| Open weights | Yes | Yes |
| Released | Jun 2026 | Dec 2025 |
| LiveBench score | 50.7 | 57.3 |
| GPQA Diamond score | 85.3 | 83.3 |
| Epoch Capabilities Index score | 146 | 144 |
| FrontierMath score | — | 2.4 |
| LMArena (Chatbot Arena) Elo score | — | 1435 |
| LMArena WebDev Elo score | — | 1435 |
FAQ
Is Nemotron 3 Ultra or GLM 4.7 cheaper?
GLM 4.7 is cheaper on output tokens ($2.20 vs $2.40 per 1M).
Which has the larger context window, Nemotron 3 Ultra or GLM 4.7?
Nemotron 3 Ultra has the larger context window (262K tokens).
Pricing is indicative — confirm with the provider before production use. Highlighted values indicate the better figure for that row.