DEPLOY

Open-source LLM · DeepSeek

DeepSeek V3

671 B parameters (37 B active per token) · 128K context · MIT · released 2024-12-26 by DeepSeek.

DeepSeek V3 needs about 1342 GB at FP16: no single GPU can hold it. Serving requires an 8-GPU HGX baseboard (INT8/INT4) or a full rack (NVL72, TPU pod) at FP16.

Where this model sits

Frontier open-weight of late 2024. MoE means only 37B parameters activate per token, so inference cost is 8B-class despite the 671B footprint on disk.

What runs it (production serving)

Datacenter GPUs and rack designs sized for DeepSeek V3. Consumer cards are not practical at this size.

Memory at different context lengths

Weights are fixed. The KV cache grows linearly with context length and can rival the weights at full window. Most guides only quote weights, then hand-wave a 15% overhead, which is only true at short context.

ContextFP16 totalINT8 totalINT4 total
4K (short chat)1342 + 15.3 = 1357.3 GB671 + 7.6 = 678.6 GB336 + 7.6 = 343.1 GB
32K (typical RAG)1342 + 122.0 = 1464.0 GB671 + 61.0 = 732.0 GB336 + 61.0 = 396.5 GB
128K (full window)1342 + 488.0 = 1830.0 GB671 + 244.0 = 915.0 GB336 + 244.0 = 579.5 GB

KV bytes per token = 2 × 61 layers × 128 KV heads × 128 head-dim × precision bytes. INT4 rows keep the KV cache at INT8 (typical serving default).

Quality benchmarks

Vendor-reported scores from the release blog or tech report. Every row cites its source.

BenchmarkScoreSourceDate
MMLU (5-shot)88.5DeepSeek V3 tech report2024-12-26
HumanEval82.6DeepSeek V3 tech report2024-12-26
MATH (0-shot)61.6DeepSeek V3 tech report2024-12-26
GPQA59.1DeepSeek V3 tech report2024-12-26

Speed and cost to serve

What cloud providers currently charge to serve DeepSeek V3. Tokens per second is what the provider publishes on their own page. Batch size, prompt length, and inference engine can shift these numbers a lot; treat as directional.

ProviderHardwareSpeed$/M in$/M outAs of
DeepSeek APIH800—$0.27$1.102025-08-01
Together AIH100 nodes—$1.25$1.252025-08-01
Fireworks AIH100 nodes—$0.90$0.902025-08-01

Cheapest option in this list for output-heavy workloads: Fireworks AI at $0.90/M out. Fastest published: not enough data.

License and variants

License
MIT
The gotcha
MIT weights, but the DeepSeek terms of use restrict certain military and surveillance applications.
Variants
Base · Chat
Context window
131,072 tokens (128K)
Architecture
MoE (671B total, 37B active per token)

Compare with

Model card: https://huggingface.co/deepseek-ai/DeepSeek-V3 · Release: DeepSeek announcement

See also: every open-source LLM we track · all AI chips · rack designs.