Google95 GB VRAM819 GB/s$4.2/hr

Qwen 3 4B on TPU v5p (per-chip, 95GB): Does It Fit, How Fast, What It Costs

Qwen 3 4B at FP16 needs 8.0 GB for its weights and 17.7 GB in total at 8K context with 8 concurrent users. That means it fits on a single TPU v5p (per-chip, 95GB). Every figure is computed by tokcalc's open-source capacity model.

Fits?
Yes, 1 GPU
17.7 GB needed / 95 GB available
Decode speed
66.54
tok/s per stream
Aggregate
798.5
tok/s @ 8 concurrent
Cost / 1M tokens
$1.46
1× TPU v5p (per-chip, 95GB)
What changes as you add TPU v5p (per-chip, 95GB)s
Tensor-parallel degree vs VRAM, throughput, and cost. Cost scales linearly with GPU count; decode throughput scales near-linearly too, so cost per token stays roughly flat.
GPUsTotal VRAMVRAM neededDecode tok/sAggregate tok/sTTFTCost / 1M tokFits?
1× TPU95 GB17.7 GB66.54798.5330.6 ms$1.46✓
2× TPU190 GB17.7 GB113.12715181.6 ms$0.8594✓
4× TPU380 GB17.7 GB226.210860105.8 ms$0.4297✓
8× TPU760 GB17.7 GB452.54344067.9 ms$0.2149✓
Concurrency vs context length
On 1× TPU v5p (per-chip, 95GB) at FP16. KV cache scales linearly with context, so this is the number of users your GPUs can actually serve.
ContextConcurrent usersKV per requestFits?
4K144618.5 MB✓ fits
8K721.21 GB✓ fits
33K184.83 GB✓ fits
66K99.66 GB✓ fits
131K419.33 GB✓ fits
Other GPUs that run Qwen 3 4B
Same workload, same quantization — ranked by value (tokens per dollar-hour).
GPUGPUsAggregate tok/sCost / 1M tok
RTX 3090 24GB1×912.6$0.0761
RTX 5090 32GB1×1747$0.0795
RTX 4090 24GB1×982.8$0.1131
AMD MI300X 192GB1×5168$0.1344
V100 SXM2 32GB1×877.5$0.1583
A100 40GB SXM41×1511$0.1654
Frequently asked questions

Can Qwen 3 4B run on TPU v5p (per-chip, 95GB)?

Yes. Qwen 3 4B at FP16 needs 17.7 GB including the KV cache, and TPU v5p (per-chip, 95GB) has 95 GB — so it fits on a single GPU.

How fast is Qwen 3 4B on TPU v5p (per-chip, 95GB)?

66.54 tokens/sec per stream and 798.5 tok/s aggregate with continuous batching at 8 concurrent users. Time to first token is 330.6 ms, and inter-token latency is 15.0 ms.

What does Qwen 3 4B on TPU v5p (per-chip, 95GB) cost per million tokens?

$1.46 per 1M output tokens at 1× TPU v5p (per-chip, 95GB) ($4.20/hr). That is the marginal hardware cost — add idle capacity, ops, and egress for the real figure.

How much VRAM does Qwen 3 4B need on TPU v5p (per-chip, 95GB)?

8.0 GB for the weights at FP16, plus 9.7 GB of KV cache for 8 concurrent requests at 8K context — 17.7 GB total. Every extra concurrent user at long context costs more, so leave headroom.

Is TPU v5p (per-chip, 95GB) the right choice for Qwen 3 4B?

It works. Compare $1.46 per 1M tokens against the alternatives below — the cheapest option that fits the same workload is 2× RTX 3080 10GB at $0.0441, and the fastest is 1× Cerebras CS-3 (WSE-3) at 19500 tok/s.

Adjust the workload yourself

Same formulas, your context length, concurrency, and quantisation.

Open in calculator