Llama 3.1 Nemotron 70B Instruct
Oct 2024
Llama 3.1 Nemotron 70B InstructCompare inon
Benchmarks of Llama 3.1 Nemotron 70B Instruct11
BenchmarkByScoreRank
No benchmark matches
Ranking?
Alternatives?
ModelScoreCostPer taskWhat changes
CheaperMiMo v2.6 ProunstatedOpenwins 75%-55%$0.43 / $0.87Pass back reasoning_content+3
BetterGPT-5 Minihighwins 59%-43%$0.053CheaperMiMo v2.5 ProunstatedOpenwins 72%-55%$0.43 / $0.87CheaperMiMo v2.6 FlashunstatedOpenwins 71%-85%$0.14 / $0.28Pass back reasoning_content+3
CheaperGLM 5.3 FlashunstatedOpenwins 71%-80%$0.15 / $0.50Defaults to max · Self-hosted+4ModelScoreCostPer taskWhat changes
MiMo v2.6 ProunstatedOpenwins 75%-55%$0.43 / $0.87Pass back reasoning_content+3MiMo v2.5 ProunstatedOpenwins 72%-55%$0.43 / $0.87MiMo v2.6 FlashunstatedOpenwins 71%-85%$0.14 / $0.28Pass back reasoning_content+3GLM 5.3 FlashunstatedOpenwins 71%-80%$0.15 / $0.50Defaults to max · Self-hosted+4
Gemini 3 Flash Previewunstatedwins 71%-6%$0.50 / $3ModelScoreCostPer taskWhat changes
GPT-5 Minihighwins 59%-43%$0.053GPT-6 Lunamaxwins 61%-83%$0.068Function calling only with none · Chat Completions+3
Claude Haiku 5.5highwins 63%-83%$0.079Remove sampling parameters+5GPT-5.6 Lunaxhighwins 67%-63%$0.085GPT-5.6 Terraxhighwins 68%+275%$0.63Prompts with >272K input tokensModelScoreCostPer taskWhat changes
MiMo v2.6 ProunstatedOpenwins 75%-55%$0.43 / $0.87Pass back reasoning_content+3GLM 5.3maxOpenwins 72%+79%$2.01Reasoning cannot be disabled+3
Kimi K3maxOpenwins 72%+400%$2.00Sampling parameters are fixed+3MiMo v2.5 ProunstatedOpenwins 72%-55%$0.43 / $0.87GLM 5.1unstatedOpenwins 72%+79%$1.40 / $4.40ModelScoreCostPer taskWhat changes
NVIDIA Nemotron 3 Ultra 550B A55B NVFP4Openwins 69%——NVIDIA Nemotron 3.5 Super VL 120B A12B NVFP4Openwins 66%—$0.08 / $0.45Nemotron 3 Nano 30B A3BunstatedOpenwins 59%—$0.06 / $0.24NVIDIA Nemotron 3.5 Lightning 30B A3B NVFP4Openwins 58%——Llama Nemotron Super 49B v1.5unstatedOpenwins 56%—$0.40 / $0.40Context: 131K → 128KLlama 3 3 Nemotron Super 49B v1Openwins 56%——No strict JSONNemotron 4 340B InstructOpenwins 39%—$4.20 / $4.20Context: 131K → 4KNVIDIA Nemotron Labs Dory 1.0 4BOpenn/a——All NVIDIA models →Run it yourself
Fits onhardware needs not verified yet
LicenceLlama 3.1 licencecommercial use with conditions
Benchmarks
all categoriesChat & Knowledge
Arena Text · overall1,299#157
Arena Text · Chinese1,263.1#158
Arena Text · English1,314.8#156
Arena Text · instruction following1,251.7#163
Arena Text · non-English1,244.9#166
Arena Text · Russian1,226.7#171
Arena Text · overall · October 22, 20261,282.6#155
Arena Text · overall · October 15, 20261,282.6#149
Writing & Language
Arena Text · creative writing1,276#145
FAQ
How much does Llama 3.1 Nemotron 70B Instruct cost?›
$1.20 per 1M input tokens and $1.20 per 1M output tokens (cheapest host: DeepInfra).
Is Llama 3.1 Nemotron 70B Instruct free to use?›
There is no free API; the cheapest paid price is $1.20 per 1M input tokens. The weights are open, so you can run it yourself.
How good is Llama 3.1 Nemotron 70B Instruct, and what is it best at?›
#156 of 205 in Chat & Knowledge, #144 of 197 in Writing & Language.
Is Llama 3.1 Nemotron 70B Instruct better than Granite 4.1 8B?›
On Arena Text · English, Llama 3.1 Nemotron 70B Instruct scores 1,314.8 vs 1,315.3 — about the same.
How can I use Llama 3.1 Nemotron 70B Instruct?›
Through the API of DeepInfra.
What is the context window of Llama 3.1 Nemotron 70B Instruct?›
131K tokens.
When was Llama 3.1 Nemotron 70B Instruct released? Is there a newer version?›
Released 2024-10-12. It is the current version in its line.
Is Llama 3.1 Nemotron 70B Instruct open source?›
Yes, the weights are open (llama3.1 licence).
Can I run Llama 3.1 Nemotron 70B Instruct locally?›
The weights are open and can be self-hosted; hardware needs are not confirmed yet.
Can Llama 3.1 Nemotron 70B Instruct read images?›
No, it accepts text only.
What is Llama 3.1 Nemotron 70B Instruct?›
Llama 3.1 Nemotron 70B Instruct is an open-weights language model from NVIDIA, released Oct 2024.
What are the best alternatives to Llama 3.1 Nemotron 70B Instruct?›
MiMo v2.6 Pro (cheaper), GPT-5 Mini (better).