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. | 24B | 49B |
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. | 32k | 128k |
| 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. | $0.04DeepInfra | — |
Cheapest outputLowest output rate across every provider on OpenRouter, the lab included. May come from a different provider than the cheapest input. | $0.064DeepInfra | — |
| $14.00DeepInfra · fp8 | — | |
| Benchmarks | ||
MMLU-ProKnowledge — accuracy on the ten-option MMLU-Pro benchmark | 66.3% | 79.53% |
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. | — | 73.58% |
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. | 84.8% | — |
IFEvalAgentic tasks — following verifiable instructions in generated responses | 82.9% | — |
Humanity's Last Exam · text only, no toolsMultidisciplinary reasoning — accuracy on the 2,158 text-only questions without tools | — | 7.64% |
MATHReasoning — accuracy on the original competition mathematics test set | 70.6% | — |
MATH-500Reasoning — accuracy on the 500-problem MATH subset | — | 97.4% |
AIME 2024Reasoning — single-answer accuracy on the 2024 invitational mathematics exam without tools | — | 87.5% |
AIME 2025Reasoning — single-answer accuracy on the 2025 invitational mathematics exam without tools | — | 82.71% |
GPQA DiamondScience — Graduate-level science questions in biology, physics, and chemistry — hard enough that subject-matter PhDs score around 65%. Higher is better. | — | 71.97% |
| Overview | ||
| Company | Mistral | NVIDIA |
| Release date | Jan 30 2025 | Jul 25 2025 |
| Access | Open Weight | Open Weight |
| Model details | View model | View model |
Other comparisons
Mistral Small 3vsClaude Fable 5.2Llama Nemotron Super 49B v1.5vsClaude Fable 5.2Mistral Small 3vsGPT-6.2 AstraLlama Nemotron Super 49B v1.5vsGPT-6.2 AstraMistral Small 3vsGemini 4 NeonLlama Nemotron Super 49B v1.5vsGemini 4 NeonMistral Small 3vsMuse EmberLlama Nemotron Super 49B v1.5vsMuse EmberMistral Small 3vsGrok 4.7Llama Nemotron Super 49B v1.5vsGrok 4.7Mistral Small 3vsDeepSeek-V4.1-Flash-1017Llama Nemotron Super 49B v1.5vsDeepSeek-V4.1-Flash-1017Frequently asked questions
Llama Nemotron Super 49B v1.5 leads Mistral Small 3 on 1 of the 1 benchmark they both report (MMLU-Pro). Mistral Small 3 shipped 176 days before Llama Nemotron Super 49B v1.5, so benchmark comparisons should account for the intervening progress.
Mistral Small 3 has 24B parameters, while Llama Nemotron Super 49B v1.5 has 49B. Context windows are 32k (Mistral Small 3) vs 128k (Llama Nemotron Super 49B v1.5).
On MMLU-Pro, Llama Nemotron Super 49B v1.5 leads at 79.53% vs Mistral Small 3 at 66.3%.