- Intelligence
- 22.9
- Rank 203 of 744 measured models
- Coding
- 49.3
- Rank 125 of 298 measured models
- Output speed
- 155 t/s
- Rank 78 of 403 measured models
- Price/1M
- $1.08
- Rank 387 of 659 measured models
- Context window
- 262K
- Rank 458 of 1199 measured models
- Cost per task
- $0.598
- Rank 111 of 193 measured models
Nemotron 3 Ultra 550B A55B (Reasoning) scores 22.9 on the Artificial Analysis Intelligence Index, ranking 203rd of 744 measured models. The closest models on this index are GPT-5 (23.0), Qwen3.5 27B (22.9), and MiniMax-M2.5 (22.8). At $1.08 per million tokens (3:1 input/output blended price), it is cheaper than GPT-5 ($3.44), more expensive than Qwen3.5 27B ($0.825) and MiniMax-M2.5 ($0.525). With an output speed of 155 t/s, it is faster than GPT-5 (102 t/s), Qwen3.5 27B (76.5 t/s), and MiniMax-M2.5 (66.5 t/s).
Models at a similar level
The closest models on the Intelligence Index, one configuration per family.
- GPT-5
OpenAI · $3.44 · 102 t/s
23.0+0.1$3.44102 t/sCompare - Qwen3.5 27B
Alibaba · $0.825 · 76.5 t/s
22.9$0.82576.5 t/sCompare - MiniMax-M2.5
MiniMax · $0.525 · 66.5 t/s
22.8−0.1$0.52566.5 t/sCompare - GPT-5.5 Instant (May 2026)
OpenAI · $11.25 · —
22.7−0.2$11.25—Compare
