NVIDIA48 GB VRAM864 GB/s$0.7/hr

Codestral 25.08 on L40S 48GB: Does It Fit, How Fast, What It Costs

Codestral 25.08 at FP8 (E4M3) needs 22.0 GB for its weights and 37.0 GB in total at 8K context with 8 concurrent users. That means it fits on a single L40S 48GB. Every figure is computed by tokcalc's open-source capacity model.

Fits?
Yes, 1 GPU
37.0 GB needed / 48 GB available
Decode speed
38.29
tok/s per stream
Aggregate
459.5
tok/s @ 8 concurrent
Cost / 1M tokens
$0.4232
1× L40S 48GB
What changes as you add L40S 48GBs
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× L40S48 GB37.0 GB38.29459.57.93 s$0.4232✓
2× L40S96 GB37.0 GB65.0915623.98 s$0.2489✓
4× L40S192 GB37.0 GB130.262492.01 s$0.1245✓
8× L40S384 GB37.0 GB260.4249961.02 s$0.0622✓
Concurrency vs context length
On 1× L40S 48GB at FP8 (E4M3). KV cache scales linearly with context, so this is the number of users your GPUs can actually serve.
ContextConcurrent usersKV per requestFits?
4K27962.1 MB✓ fits
8K131.88 GB✓ fits
33K37.52 GB✓ fits
66K115.03 GB✓ fits
131K030.06 GB✗ does not fit
Other GPUs that run Codestral 25.08
Same workload, same quantization — ranked by value (tokens per dollar-hour).
GPUGPUsAggregate tok/sCost / 1M tok
RTX 5090 32GB2×3240$0.0857
AMD MI300X 192GB1×2819$0.2464
A100 40GB SXM41×824.3$0.3033
AMD MI325X 288GB1×3191$0.3047
A100 80GB SXM41×1064$0.3134
Intel Gaudi 3 128GB1×1755$0.3166
Frequently asked questions

Can Codestral 25.08 run on L40S 48GB?

Yes. Codestral 25.08 at FP8 (E4M3) needs 37.0 GB including the KV cache, and L40S 48GB has 48 GB — so it fits on a single GPU.

How fast is Codestral 25.08 on L40S 48GB?

38.29 tokens/sec per stream and 459.5 tok/s aggregate with continuous batching at 8 concurrent users. Time to first token is 7.93 s, and inter-token latency is 26.1 ms.

What does Codestral 25.08 on L40S 48GB cost per million tokens?

$0.4232 per 1M output tokens at 1× L40S 48GB ($0.7000/hr). That is the marginal hardware cost — add idle capacity, ops, and egress for the real figure.

How much VRAM does Codestral 25.08 need on L40S 48GB?

22.0 GB for the weights at FP8 (E4M3), plus 15.0 GB of KV cache for 8 concurrent requests at 8K context — 37.0 GB total. Every extra concurrent user at long context costs more, so leave headroom.

Is L40S 48GB the right choice for Codestral 25.08?

It works. Compare $0.4232 per 1M tokens against the alternatives below — the cheapest option that fits the same workload is 4× RTX 3080 10GB at $0.0404, and the fastest is 1× Cerebras CS-3 (WSE-3) at 10636 tok/s.

Adjust the workload yourself

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

Open in calculator