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 | 1B/3B/11B/90B |
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 | 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. | — | $0.027Cloudflare |
Cheapest outputLowest output rate across every provider on OpenRouter, the lab included. May come from a different provider than the cheapest input. | — | $0.201Cloudflare |
| — | $15.45Cloudflare | |
| These models have no shared benchmark scores. | ||
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% | — |
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% | — |
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% | — |
MATHReasoning — accuracy on the original competition mathematics test set | — | 68% |
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. | — | 86% |
MGSMMultilingual math — solving grade-school word problems translated into multiple languages | — | 86.9% |
MathVistaMultimodal reasoning — mathematical reasoning about visual information | — | 57.3% |
MMMUMultimodal — Tests the AI on understanding images and text together across many college subjects. Higher is better. | — | 60.3% |
| Overview | ||
| Company | DeepSeek | Meta |
| Release date | Jul 24 2024 | Sep 25 2024 |
| Access | Open Weight | Open Weight |
| Model details | View model | View model |
Other comparisons
DeepSeek Coder V2-0724vsClaude Fable 5.2LLaMA 3.2vsClaude Fable 5.2DeepSeek Coder V2-0724vsGPT-6.2 AstraLLaMA 3.2vsGPT-6.2 AstraDeepSeek Coder V2-0724vsGemini 4 HydrogenLLaMA 3.2vsGemini 4 HydrogenDeepSeek Coder V2-0724vsGrok 4.7LLaMA 3.2vsGrok 4.7DeepSeek Coder V2-0724vsMistral Large 4LLaMA 3.2vsMistral Large 4DeepSeek Coder V2-0724vsKimi K3.1 ProLLaMA 3.2vsKimi K3.1 ProFrequently asked questions
DeepSeek Coder V2-0724 and LLaMA 3.2 don't publish scores on any of the same benchmarks, so there's no direct head-to-head comparison. DeepSeek Coder V2-0724 shipped 63 days before LLaMA 3.2, so benchmark comparisons should account for the intervening progress.
DeepSeek Coder V2-0724 has 236B parameters, while LLaMA 3.2 has 1B/3B/11B/90B. Context windows are 128k (DeepSeek Coder V2-0724) vs 128k (LLaMA 3.2).
Direct benchmark comparisons are unavailable — DeepSeek Coder V2-0724 and LLaMA 3.2 don't publish scores on any of the same benchmarks.