Compare AI models
| Specifications | ||
ParametersThe learned values that shape how a model responds. B means billion and T means trillion. For mixture-of-experts models, the total count is larger than the number active for each token. Parameter count alone does not measure quality. | 120B | 6B |
Context windowHow much information the model can work with in one request, measured in tokens — pieces of words or other data. This includes your question, attachments, conversation and reply. Limits can vary by model variant and provider. | 1M | 32k |
| API pricingUSD per 1M tokens · lower wins | ||
Cheapest inputLowest input rate across every provider on OpenRouter, the lab included. The cheapest endpoint may run a quantised build or a shorter context — see "Available from" on the model page. | $0.00Nvidia | — |
Cheapest outputLowest output rate across every provider on OpenRouter, the lab included. May come from a different provider than the cheapest input. | $0.00Nvidia | — |
| $33.00Io Net · nvfp4 | — | |
| These models have no shared benchmark scores. | ||
SWE-Bench ProAgentic coding — Can the AI fix real bugs in real software? It's handed actual problems from open-source projects and has to write code that genuinely solves them. Higher is better. | 51.78% | — |
SWE-Bench VerifiedCoding — Real coding tasks pulled from open-source projects — the AI has to find and fix actual bugs. A human-checked version of the original SWE-Bench. Higher is better. | 71.1% | — |
SWE-Bench MultilingualMultilingual coding — Like SWE-Bench, but the coding problems span many programming languages, not just one. Tests how broadly the AI can code. Higher is better. | 67.72% | — |
LiveCodeBenchCompetitive coding — Coding problems published so recently the AI can't have seen them in training — a contamination-free test of raw programming skill. Higher is better. | 88.22% | — |
HumanEvalFunction synthesis — 164 small Python problems: the AI is given a function's description and has to write the working function. This was the coding benchmark of the GPT-3.5 and GPT-4 era, before the field moved to fixing real bugs in real repositories. Higher is better. | — | 9.8% |
Terminal-Bench 4.0Agentic terminal coding — Can the AI work in a command-line terminal — running commands and finishing technical setup tasks the way a developer would? Version 4.0 recalibrated how much time, CPU and memory each task gets, removed eight tasks and fixed nineteen, so fewer runs fail for reasons that have nothing to do with the model. Scores are not comparable with earlier versions. Higher is better. | 0.3% | — |
Terminal-Bench 2.1Agentic terminal coding — Can the AI work in a command-line terminal — running commands and finishing technical setup tasks the way a developer would? Higher is better. | 66.7% | — |
BrowseCompWeb browsing — Can the AI browse the web and track down hard-to-find answers? Higher is better. | 78% | — |
Humanity's Last Exam · with toolsMultidisciplinary reasoning — Humanity's Last Exam — extremely hard expert questions across many subjects. “With tools” means the AI is allowed to search the web or run code while answering. Higher is better. | 29.2% | — |
MATHReasoning — accuracy on the original competition mathematics test set | — | 6.9% |
HellaSwagReasoning — commonsense completion of short scenarios | — | 57% |
GPQA DiamondScience — Graduate-level science questions in biology, physics, and chemistry — hard enough that subject-matter PhDs score around 65%. Higher is better. | 89.52% | — |
IFBenchInstruction following — Tests whether the AI can follow detailed instructions and satisfy multiple constraints at once. Higher is better. | 70.86% | — |
MMLUGeneral knowledge — A 57-subject multiple-choice exam — history, law, medicine, maths — that was the standard measure of how much a model knows from 2020 until roughly 2024, when frontier scores crowded into the high 80s and labs moved on to harder tests. The scores here were published years apart under different testing setups, so read them as a historical record rather than a like-for-like ranking. Higher is better. | — | 45.2% |
GSM8KGrade-school math — Grade-school maths word problems that take a few steps of arithmetic to work through. It separated the models of 2022 and 2023 sharply, then saturated. One caveat on the historical numbers: OpenAI included part of the GSM8K training set in GPT-4's pre-training mix, so GPT-4's score is not a clean few-shot result. Higher is better. | — | 25.9% |
| Overview | ||
| Company | NVIDIA | Z.ai |
| Release date | Oct 15 2026 | Jun 25 2023 |
| Access | Open Weight | Open Weight |
| Model details | View model | View model |
Other comparisons
Nemotron 3.5 Super VLvsClaude Fable 5.2ChatGLM2-6BvsClaude Fable 5.2Nemotron 3.5 Super VLvsGPT-6.2 AstraChatGLM2-6BvsGPT-6.2 AstraNemotron 3.5 Super VLvsGemini 4 HydrogenChatGLM2-6BvsGemini 4 HydrogenNemotron 3.5 Super VLvsMuse EmberChatGLM2-6BvsMuse EmberNemotron 3.5 Super VLvsGrok 4.7ChatGLM2-6BvsGrok 4.7Nemotron 3.5 Super VLvsDeepSeek-V4.1-Flash-1017ChatGLM2-6BvsDeepSeek-V4.1-Flash-1017Frequently asked questions
Nemotron 3.5 Super VL and ChatGLM2-6B don't publish scores on any of the same benchmarks, so there's no direct head-to-head comparison. ChatGLM2-6B shipped 1208 days before Nemotron 3.5 Super VL, so benchmark comparisons should account for the intervening progress.
Nemotron 3.5 Super VL has 120B parameters, while ChatGLM2-6B has 6B. Context windows are 1M (Nemotron 3.5 Super VL) vs 32k (ChatGLM2-6B).
Direct benchmark comparisons are unavailable — Nemotron 3.5 Super VL and ChatGLM2-6B don't publish scores on any of the same benchmarks.