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. | 236B | 400B |
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 | 1M |
| 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. | — | $0.35Google |
Cheapest outputLowest output rate across every provider on OpenRouter, the lab included. May come from a different provider than the cheapest input. | — | $1.00Parasail |
| — | $120Parasail · fp8 | |
| 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. | 39.7% | 43.4% |
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. | — | 29% |
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. | 19% | — |
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. | 87.2% | — |
GPQA DiamondScience — Graduate-level science questions in biology, physics, and chemistry — hard enough that subject-matter PhDs score around 65%. Higher is better. | — | 69.8% |
MMLU-ProKnowledge — accuracy on the ten-option MMLU-Pro benchmark | — | 80.5% |
MGSMMultilingual math — solving grade-school word problems translated into multiple languages | — | 92.3% |
MathVistaMultimodal reasoning — mathematical reasoning about visual information | — | 73.7% |
MMMU-Pro · standard and vision meanMultimodal reasoning — average across the standard and vision sets | — | 59.6% |
MMMUMultimodal — Tests the AI on understanding images and text together across many college subjects. Higher is better. | — | 73.4% |
| Overview | ||
| Company | DeepSeek | Meta |
| Release date | Jul 24 2024 | Apr 5 2025 |
| Access | Open Weight | Open Weight |
| Model details | View model | View model |
Other comparisons
DeepSeek Coder V2-0724vsClaude Fable 5.2LLaMA 4 MaverickvsClaude Fable 5.2DeepSeek Coder V2-0724vsGPT-6.2 AstraLLaMA 4 MaverickvsGPT-6.2 AstraDeepSeek Coder V2-0724vsGemini 4 HydrogenLLaMA 4 MaverickvsGemini 4 HydrogenDeepSeek Coder V2-0724vsGrok 4.7LLaMA 4 MaverickvsGrok 4.7DeepSeek Coder V2-0724vsMistral Large 4LLaMA 4 MaverickvsMistral Large 4DeepSeek Coder V2-0724vsKimi K3.1 ProLLaMA 4 MaverickvsKimi K3.1 ProFrequently asked questions
LLaMA 4 Maverick leads DeepSeek Coder V2-0724 on 1 of the 1 benchmark they both report (LiveCodeBench). DeepSeek Coder V2-0724 shipped 255 days before LLaMA 4 Maverick, so benchmark comparisons should account for the intervening progress.
DeepSeek Coder V2-0724 has 236B parameters, while LLaMA 4 Maverick has 400B. Context windows are 128k (DeepSeek Coder V2-0724) vs 1M (LLaMA 4 Maverick).
On LiveCodeBench, LLaMA 4 Maverick leads at 43.4% vs DeepSeek Coder V2-0724 at 39.7%.