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. | 7B/13B/33B/65B | — |
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. | 2k | — |
| Benchmarks | ||
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. | 23.7% | 85.7% |
MATHReasoning — accuracy on the original competition mathematics test set | 10.6% | 73% |
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. | 63.4% | 86.2% |
ARC-ChallengeReasoning — multiple-choice science questions from the AI2 Reasoning Challenge; unrelated to ARC-AGI | 56% | — |
HellaSwagReasoning — commonsense completion of short scenarios | 84.2% | — |
WinoGrandeReasoning — resolving ambiguous pronouns in context | 77% | — |
MMLU-ProKnowledge — accuracy on the ten-option MMLU-Pro benchmark | — | 72% |
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. | 50.9% | — |
MathVistaMultimodal reasoning — mathematical reasoning about visual information | — | 68.1% |
MMMUMultimodal — Tests the AI on understanding images and text together across many college subjects. Higher is better. | — | 63.2% |
| Overview | ||
| Company | Meta | SpaceXSI |
| Release date | Feb 24 2023 | Aug 13 2024 |
| Access | Open Weight | Closed |
| Model details | View model | View model |
Other comparisons
LLaMA 1vsClaude Fable 5.2Grok‑2 minivsClaude Fable 5.2LLaMA 1vsGPT-6.1 TerraGrok‑2 minivsGPT-6.1 TerraLLaMA 1vsGemini 4 HydrogenGrok‑2 minivsGemini 4 HydrogenLLaMA 1vsDeepSeek-V4.1-Flash-1017Grok‑2 minivsDeepSeek-V4.1-Flash-1017LLaMA 1vsMistral Large 4Grok‑2 minivsMistral Large 4LLaMA 1vsKimi K3.1 ProGrok‑2 minivsKimi K3.1 ProFrequently asked questions
Grok‑2 mini leads LLaMA 1 on 3 of the 3 benchmarks they both report (HumanEval, MATH, MMLU). LLaMA 1 shipped 536 days before Grok‑2 mini, so benchmark comparisons should account for the intervening progress.
LLaMA 1 is open weight, while Grok‑2 mini is closed.
On HumanEval, Grok‑2 mini leads at 85.7% vs LLaMA 1 at 23.7%. On MATH, Grok‑2 mini leads at 73% vs LLaMA 1 at 10.6%. On MMLU, Grok‑2 mini leads at 86.2% vs LLaMA 1 at 63.4%.