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. | 4B | 0.6B/1.7B/4B/8B/14B/30B/32B/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. | 256k | — |
| 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. | — | $0.064DeepInfra |
Cheapest outputLowest output rate across every provider on OpenRouter, the lab included. May come from a different provider than the cheapest input. | — | $0.192DeepInfra |
| — | $30.00DeepInfra · fp8 | |
| Benchmarks | ||
MATH-500Reasoning — accuracy on the 500-problem MATH subset | 95.4% | 98% |
AIME 2025Reasoning — single-answer accuracy on the 2025 invitational mathematics exam without tools | 78.5% | 81.5% |
GPQA DiamondScience — Graduate-level science questions in biology, physics, and chemistry — hard enough that subject-matter PhDs score around 65%. Higher is better. | 53.2% | 71.1% |
BullshitBench v2Nonsense detection — Given a confidently-worded but nonsensical prompt, does the AI spot that it makes no sense and push back — instead of playing along and inventing an answer? The score is how often it clearly called out the nonsense. Higher is better. | — | 6% |
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. | — | 70.7% |
IFEvalAgentic tasks — following verifiable instructions in generated responses | — | 83.4% |
τ²-bench · retailAgentic tasks — task success in the retail domain of the second-generation customer-service evaluation | 39.8% | — |
τ²-bench · airlineAgentic tasks — task success in the airline domain of the second-generation customer-service evaluation | 33.3% | — |
τ²-bench · telecomAgentic tasks — task success in the telecom domain of the second-generation customer-service evaluation | 33% | — |
AIME 2024Reasoning — single-answer accuracy on the 2024 invitational mathematics exam without tools | — | 85.7% |
| Overview | ||
| Company | NVIDIA | Qwen |
| Release date | Mar 16 2026 | Apr 29 2025 |
| Access | Open Weight | Open Weight |
| Model details | View model | View model |
Other comparisons
Nemotron 3 Nano 4BvsClaude Fable 5.2Qwen3vsClaude Fable 5.2Nemotron 3 Nano 4BvsGPT-6.2 AstraQwen3vsGPT-6.2 AstraNemotron 3 Nano 4BvsGemini 4 NeonQwen3vsGemini 4 NeonNemotron 3 Nano 4BvsMuse EmberQwen3vsMuse EmberNemotron 3 Nano 4BvsGrok 4.7Qwen3vsGrok 4.7Nemotron 3 Nano 4BvsDeepSeek-V4.1-Flash-1017Qwen3vsDeepSeek-V4.1-Flash-1017Frequently asked questions
Qwen3 leads Nemotron 3 Nano 4B on 3 of the 3 benchmarks they both report (MATH-500, AIME 2025, GPQA Diamond). Qwen3 shipped 321 days before Nemotron 3 Nano 4B, so benchmark comparisons should account for the intervening progress.
Nemotron 3 Nano 4B has 4B parameters, while Qwen3 has 0.6B/1.7B/4B/8B/14B/30B/32B/235B.
On MATH-500, Qwen3 leads at 98% vs Nemotron 3 Nano 4B at 95.4%. On AIME 2025, Qwen3 leads at 81.5% vs Nemotron 3 Nano 4B at 78.5%. On GPQA Diamond, Qwen3 leads at 71.1% vs Nemotron 3 Nano 4B at 53.2%.