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. | 253B | 4B/30B/235B |
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. | 128k | 256k |
| Benchmarks | ||
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. | 66.31% | 74.1% |
AIME 2025Reasoning — single-answer accuracy on the 2025 invitational mathematics exam without tools | 72.5% | 92.3% |
GPQA DiamondScience — Graduate-level science questions in biology, physics, and chemistry — hard enough that subject-matter PhDs score around 65%. Higher is better. | 76.01% | 81.1% |
IFEvalAgentic tasks — following verifiable instructions in generated responses | — | 87.8% |
Humanity's Last Exam · text only, no toolsMultidisciplinary reasoning — accuracy on the 2,158 text-only questions without tools | — | 18.2% |
MATH-500Reasoning — accuracy on the 500-problem MATH subset | 97% | — |
AIME 2024Reasoning — single-answer accuracy on the 2024 invitational mathematics exam without tools | 80.8% | — |
MMLU-ProKnowledge — accuracy on the ten-option MMLU-Pro benchmark | — | 84.4% |
| Overview | ||
| Company | NVIDIA | Qwen |
| Release date | Apr 7 2025 | Jul 21 2025 |
| Access | Open Weight | Open Weight |
| Model details | View model | View model |
Other comparisons
Llama Nemotron Ultra 253BvsClaude Fable 5.2Qwen3-2507vsClaude Fable 5.2Llama Nemotron Ultra 253BvsGPT-6.2 AstraQwen3-2507vsGPT-6.2 AstraLlama Nemotron Ultra 253BvsGemini 4 NeonQwen3-2507vsGemini 4 NeonLlama Nemotron Ultra 253BvsMuse EmberQwen3-2507vsMuse EmberLlama Nemotron Ultra 253BvsGrok 4.7Qwen3-2507vsGrok 4.7Llama Nemotron Ultra 253BvsDeepSeek-V4.1-Flash-1017Qwen3-2507vsDeepSeek-V4.1-Flash-1017Frequently asked questions
Qwen3-2507 leads Llama Nemotron Ultra 253B on 3 of the 3 benchmarks they both report (LiveCodeBench, AIME 2025, GPQA Diamond). Llama Nemotron Ultra 253B shipped 105 days before Qwen3-2507, so benchmark comparisons should account for the intervening progress.
Llama Nemotron Ultra 253B has 253B parameters, while Qwen3-2507 has 4B/30B/235B. Context windows are 128k (Llama Nemotron Ultra 253B) vs 256k (Qwen3-2507).
On LiveCodeBench, Qwen3-2507 leads at 74.1% vs Llama Nemotron Ultra 253B at 66.31%. On AIME 2025, Qwen3-2507 leads at 92.3% vs Llama Nemotron Ultra 253B at 72.5%. On GPQA Diamond, Qwen3-2507 leads at 81.1% vs Llama Nemotron Ultra 253B at 76.01%.