TextDense14B params131K max context

Qwen 2.5 14B VRAM Requirements: How Many GPUs & What It Costs

Qwen 2.5 14B needs 28.0 GB of VRAM for its weights at FP16 — 28.8 GB once a KV cache is loaded. Below is the full breakdown: which GPUs fit, how many you need, what it serves, and what it costs per million tokens. Every figure comes from tokcalc's open-source capacity model.

Weights (FP16)
28.0 GB
28.8 GB with KV cache
Fits on 1 GPU?
Yes
L40S 48GB or larger
Cheapest config
$0.0386
8× RTX 3080 10GB / 1M tokens
Fastest config
11517
tok/s on 8× RTX 3080 10GB
VRAM requirement by quantization
Weight footprint = params × bytes per param. The “with KV cache” column adds one request at 4K context.
QuantizationWeightsWith KV cacheSmallest single GPU
FP16recommended28.0 GB28.8 GBV100 SXM2 32GB (32 GB)
FP8 (E4M3)14.0 GB14.8 GBT4 16GB (16 GB)
AWQ 4-bit7.7 GB8.5 GBRTX 3080 10GB (10 GB)
Which GPU should you use?
Reference workload: 8K context, 8 concurrent users, continuous batching, FP16. Ranked by fitting first, then by tokens per dollar-hour.
GPUGPUsTotal VRAMDecode tok/sAggregate tok/sCost / 1M tokens
RTX 3080 10GB8×80 GB120.011517$0.0386
RTX 4080 16GB4×64 GB56.592716$0.1227
RTX 3090 24GB2×48 GB36.94886.5$0.1567
RTX 5090 32GB2×64 GB70.721697$0.1637
T4 16GB4×64 GB25.261212$0.1833
RTX A4000 16GB4×64 GB35.361697$0.1964
RTX 4090 24GB2×48 GB39.78954.7$0.2328
V100 SXM2 32GB2×64 GB35.52852.4$0.3259
A100 40GB SXM42×80 GB61.171468$0.3406
A10G 24GB2×48 GB23.68568.3$0.3519
AMD MI300X 192GB1×192 GB123.01476$0.4704
TPU v6e Trillium (32GB)2×64 GB32.32775.7$0.5730

RTX 3080 10GB has no NVLink, so 8-way tensor parallelism runs over PCIe. tokcalc applies a flat 0.85× multi-GPU efficiency factor for this, which is realistic for NVLink-equipped datacenter cards but optimistic for PCIe scale-out — treat the throughput and cost figures for this configuration as an upper bound, and measure before committing.

How concurrency changes with context length
On 8× RTX 3080 10GB at FP16. KV cache scales linearly with context, so long contexts cut the number of users you can serve from the same GPUs.
ContextConcurrent usersKV per requestFits?
4K64824.6 MB✓ fits
8K321.61 GB✓ fits
33K86.44 GB✓ fits
66K412.88 GB✓ fits
131K225.77 GB✓ fits
Frequently asked questions

Can Qwen 2.5 14B run on a single GPU?

Yes. At FP16 the model needs 28.8 GB of VRAM, so it fits on a single L40S 48GB or anything with more VRAM. At full FP16 it needs 28.8 GB, which fits on one V100 SXM2 32GB or larger.

How many GPUs does Qwen 2.5 14B need?

At FP16 with 8,192-token context and 8 concurrent users, the cheapest configuration that fits is 8× RTX 3080 10GB (80 GB total VRAM). $0.0386 per 1M output tokens.

How fast can Qwen 2.5 14B be served?

On 8× RTX 3080 10GB at FP16, the aggregate throughput is 11517 tokens/sec with continuous batching, at $0.0386 per 1M output tokens.

How much VRAM does Qwen 2.5 14B need?

Weights alone: 28.0 GB at FP16, 28.0 GB at FP16. Add KV cache for your context length and concurrency — every concurrent request at 8,192 tokens costs extra VRAM, so budget headroom above the weight figure.

Is it cheaper to self-host Qwen 2.5 14B than to call an API?

At 8× RTX 3080 10GB the marginal cost is $0.0386 per 1M output tokens, before you add idle capacity, ops staff, or egress. That is typically cheaper than frontier API pricing at sustained volume, but API calls win when your load is bursty or too small to keep a GPU busy. The Build vs Buy tab on tokcalc prices this against a specific provider for you.

Try Qwen 2.5 14B with your own context and concurrency

The calculator uses the exact same formulas as this page — change the workload and watch every number update.

Open in calculator

Qwen 2.5 14B on specific GPUs

Qwen 2.5 14B on RTX 5090 32GBQwen 2.5 14B on AMD MI300X 192GBQwen 2.5 14B on AMD MI325X 288GBQwen 2.5 14B on A100 80GB SXM4Qwen 2.5 14B on Intel Gaudi 3 128GBQwen 2.5 14B on H100 SXM5 80GBQwen 2.5 14B on B200 SXM 192GBQwen 2.5 14B on L40S 48GB
Model architecture
The numbers behind the KV-cache and throughput math.
Total parameters14.00B
Active parameters per token14.00B
ArchitectureDense
Attention layoutFull attention (GQA)
Transformer layers48
Hidden dimension5120
Attention heads40 query / 8 KV (GQA)
Head dimension128
Vocabulary size152,064
Max context (catalog)131K
KV cache per token @ 8K192.0 KB

Every layer caches keys and values for the whole context, so the KV cache grows linearly with conversation length.