Nemotron 3.5 Bolt 74B-A9B
Oct 2026
Nemotron 3.5 Bolt 74B-A9BCompare inon
Nemotron 3.5 Bolt 74B-A9B measured on 8 of 104 Chat & Knowledge benchmarks8
BenchmarkByScoreRank
Overall1
Knowledge3
Science2
Documents2
No benchmark matches
—per task?
self-host onlyprice
—output speed?
256Kcontext
Alternatives?
ModelScoreCostPer taskWhat changes
BetterGPT-6 Lunamedium×2.5—$0.017Function calling only with none · Chat Completions+3
OpenQwen3.8 Flashthinking×2.6—$0.37Memory (4-bit): 45 GB → 108 GB+6
BetterClaude Haiku 5.5low×2.1—$0.024Remove sampling parameters+5BetterGPT-5.6 Lunahigh×3.9—$0.044
BetterMiMo v2.6 FlashthinkingOpen×2.8—$0.062Memory (4-bit): 45 GB → 186 GB+4ModelScoreCostPer taskWhat changes
Nothing here yet.
ModelScoreCostPer taskWhat changes
GPT-6 Lunamedium×2.5—$0.017Function calling only with none · Chat Completions+3Claude Haiku 5.5low×2.1—$0.024Remove sampling parameters+5GPT-5.6 Lunahigh×3.9—$0.044MiMo v2.6 FlashthinkingOpen×2.8—$0.062Memory (4-bit): 45 GB → 186 GB+4TInkling SmallthinkingOpen+93%—$0.089Memory (4-bit): 45 GB → 160 GB+1ModelScoreCostPer taskWhat changes
Qwen3.8 FlashthinkingOpen×2.6—$0.37Memory (4-bit): 45 GB → 108 GB+6Nemotron 3.5 Super VLOpen×2.5—$0.19Memory (4-bit): 45 GB → 75 GBQwen3.8 27BxhighOpen+27%—$1.01max_tokens range · QwenCloud+4Nemotron 3 Super 120B A12BOpen-27%—$1.64Memory (4-bit): 45 GB → 74 GB+1
Muse Glimmer 30BhighOpen-40%—$0.058Context: 256K → 131KModelScoreCostPer taskWhat changes
Nemotron 3.5 Super VLOpen×2.5—$0.19Memory (4-bit): 45 GB → 75 GBNemotron 3 UltrathinkingOpen-27%—$0.58Memory (4-bit): 45 GB → 336 GB+1Nemotron 3 Super 120B A12BOpen-27%—$1.64Memory (4-bit): 45 GB → 74 GB+1Nemotron Cascade 2 30B A3BOpen-87%——No image inputNemotron 3.5 LightningthinkingOpen-100%—$0.079No image inputNemotron 3 Nano Omni 30B A3B ReasoningOpen-100%—$0 / $0NVIDIA Nemotron 3 Nano 4BOpen-100%——No image inputNemotron 3 Nano 30B A3Bnon-thinkingOpen-100%—$0.06 / $0.24Context: 256K → 131K+1All NVIDIA models →Run it yourself
estimateFits on✗ 8 GB (phone, laptop)✗ 16 GB✗ 24 GB GPU✓ 48 GB GPU · compressed✓ 80 GB GPU · 8-bit
Licenceopenmdw-1.1commercial use with conditions
Benchmarks
in Chat & KnowledgeAA Intelligence Index 4.3.221.4#112
AA Omniscience Index-19.2#122
AA Omniscience · Accuracy17.1#200
Humanity's Last Exam24.7#101
AA-LCR (long context) 1.169.7#126
CritPt (physics)4.3#86
GDP.pdf · AA all-pass2.8#102
FAQ
How much does Nemotron 3.5 Bolt 74B-A9B cost?›
There is no per-unit API price: self-host only.
Is Nemotron 3.5 Bolt 74B-A9B free to use?›
There is no free API. The weights are open, so you can run it yourself.
How good is Nemotron 3.5 Bolt 74B-A9B, and what is it best at?›
#101 of 334 in Chat & Knowledge, #75 of 132 in Coding, #67 of 133 in Agents & Tool Use, #102 of 133 in Vision & Documents.
Is Nemotron 3.5 Bolt 74B-A9B better than K2 Horizon 375B A23B?›
On CritPt (physics), Nemotron 3.5 Bolt 74B-A9B scores 4.3 vs 4.6 — about the same.
What is the context window of Nemotron 3.5 Bolt 74B-A9B?›
256K tokens.
When was Nemotron 3.5 Bolt 74B-A9B released? Is there a newer version?›
Released 2026-10-15. It is the current version in its line.
Is Nemotron 3.5 Bolt 74B-A9B open source?›
Yes, the weights are open (openmdw-1.1 licence).
Can I run Nemotron 3.5 Bolt 74B-A9B locally?›
The weights are open and can be self-hosted; hardware needs are not confirmed yet.
Can Nemotron 3.5 Bolt 74B-A9B read images?›
It accepts text and images as input.
What is Nemotron 3.5 Bolt 74B-A9B?›
Nemotron 3.5 Bolt 74B-A9B is an open-weights language model from NVIDIA, released Oct 2026.
What are the best alternatives to Nemotron 3.5 Bolt 74B-A9B?›
GPT-6 Luna (better), Qwen3.8 Flash (open weights).