Nemotron 4 340B Instruct
Nemotron 4 340B InstructRanking?
Alternatives?
CheaperGemini 3.8 Flashhighwins 81%-64%*$1.24Sampling parameters are ignored+1
BetterGPT-5 Minihighwins 62%-84%$0.053
OpenKimi K3maxwins 78%+43%$2.00Sampling parameters are fixed+3CheaperGemini 3.6 Flashhighwins 79%-64%*$1.60
CheaperMuse Sparkunstatedwins 79%-52%$1.25 / $4.25Only n=1 · Meta / Chat Completions+1Gemini 3.8 Flashhighwins 81%-64%*$1.24Sampling parameters are ignored+1Gemini 3.6 Flashhighwins 79%-64%*$1.60Muse Sparkunstatedwins 79%-52%$1.25 / $4.25Only n=1 · Meta / Chat Completions+1Muse Spark 1.3maxwins 78%-52%$1.60Only auto tool_choice is supported+5Gemini 3 Flash Previewunstatedwins 78%-73%$0.50 / $3GPT-5 Minihighwins 62%-84%$0.053GPT-6 Lunamaxwins 73%-95%$0.068Function calling only with none · Chat Completions+3GPT-5.6 Lunaxhighwins 73%-89%$0.085GPT-5.6 Terraxhighwins 75%+7%$0.63Prompts with >272K input tokensGPT-6.1 Solmaxwins 78%-5%$0.72Remove unsupported parameters+5Kimi K3maxOpenwins 78%+43%$2.00Sampling parameters are fixed+3
GLM 5.2maxOpenwins 78%-49%$1.47GLM 5.3maxOpenwins 78%-49%$2.01Reasoning cannot be disabled+3GLM 5unstatedOpenwins 77%-63%$1 / $3.20GLM 5.1unstatedOpenwins 77%-49%$1.40 / $4.40NVIDIA Nemotron 3.5 Super VL 120B A12B NVFP4Openwins 71%—$0.08 / $0.45NVIDIA Nemotron 3 Ultra 550B A55B NVFP4Openwins 70%——Llama Nemotron Super 49B v1.5unstatedOpenwins 60%—$0.40 / $0.40Llama 3 3 Nemotron Super 49B v1Openwins 58%——NVIDIA Nemotron 3.5 Lightning 30B A3B NVFP4Openwins 56%——Llama 3.1 Nemotron 70B InstructOpenwins 55%—$1.20 / $1.20Nemotron 3 Nano 30B A3BunstatedOpensame—$0.06 / $0.24Nemotron 3.5 Super VLOpenn/a—$0.19All NVIDIA models →Run it yourself
Benchmarks
all categoriesFAQ
How much does Nemotron 4 340B Instruct cost?›
$4.20 per 1M input tokens and $4.20 per 1M output tokens (cheapest host: DeepInfra).
Is Nemotron 4 340B Instruct free to use?›
There is no free API; the cheapest paid price is $4.20 per 1M input tokens. The weights are open, so you can run it yourself.
How good is Nemotron 4 340B Instruct, and what is it best at?›
#165 of 196 in Chat & Knowledge, #166 of 195 in Writing & Language.
Is Nemotron 4 340B Instruct better than GPT OSS 20B?›
On Arena Text · creative writing, Nemotron 4 340B Instruct scores 1,238 vs 1,240 — about the same.
How can I use Nemotron 4 340B Instruct?›
Through the API of DeepInfra.
What is the context window of Nemotron 4 340B Instruct?›
4K tokens.
When was Nemotron 4 340B Instruct released? Is there a newer version?›
Released 2024-06-13. It is the current version in its line.
Is Nemotron 4 340B Instruct open source?›
Yes, the weights are open (nvidia-open-model-license licence).
Can I run Nemotron 4 340B Instruct locally?›
The weights are open and can be self-hosted; hardware needs are not confirmed yet.
Can Nemotron 4 340B Instruct read images?›
No, it accepts text only.
What is Nemotron 4 340B Instruct?›
Nemotron 4 340B Instruct is an open-weights language model from NVIDIA, released Jun 2024.
What are the best alternatives to Nemotron 4 340B Instruct?›
Gemini 3.8 Flash (cheaper), GPT-5 Mini (better), Kimi K3 (open weights).