StackSays
Menu
Decision toolsCompareCategoriesPricing dataMethodologyBuild my stack

Serverless GPU Cloud · decision tool

Cheapest GPU cloud at your hours

Modal, Beam, RunPod, Replicate, Baseten, Fal and Lambda — H100-class hourly rates at YOUR monthly GPU-hours, serverless vs always-on decoded, prices verified.

Prices verified 2026-07-17 against official pricing pages · fit weights are our estimates (see methodology) · share the URL to share your configuration

As of Jul 2026, with 50 H100-class GPU-hours/month, a $200 budget for this slot, Modal ranks first for these inputs at 86% fit and a StackSays-estimated $197.50/month (Beam: 82% fit, $177.50/month). Fit weights are our estimates; prices trace to official pages. Change any input below and the verdict recomputes.

Your situation

The verdict updates as you move these.

Narrow pick — the runner-up is close

Modalfit 86% · est $197.50+/mo

Per-second serverless containers with the widest published GPU menu (T4 through B300)

vs Beam: fit 86% against 82%, est $197.50+ against $177.50 per month

Modalfit 86%
$197.50+/mo
Beamfit 82%
$177.50/mo
Falfit 69%
$199.50/mo
RunPod Serverlessfit 66%
$227.50/mo
Lambdafit 65%
$164.50/mo
Replicatefit 60%
$274.50/mo
Basetenfit 50%
$325.00/mo

The decision, in plain words

For a solo builder at 50 H100-class GPU-hours/month and a $200.00/mo budget for the GPU cloud slot, Modal leads at 86% fit and $197.50/mo. The two rules doing the work: budget fit (~$197.5/mo is 99% of your $200/mo budget); batteries included (Bundles per-second billing, recurring free credits — fewer services to glue at team size 1).

Beam runs close (82% fit, $177.50/mo). It takes the lead if ~$177.5/mo is 89% of your $200/mo budget matters more to you than the current weights assume.

Generated from the calculation above — every number traces to the sourced data; nothing here is written by an AI making things up.

Why Modal scores 86%

Budget fit (weight 30, our estimate) — ~$197.5/mo is 99% of your $200/mo budget+15.37
Batteries included (weight 12, our estimate) — Bundles per-second billing, recurring free credits — fewer services to glue at team size 1+12
Exit cost (weight 4, our estimate) — Standard containers move easily, but queueing/autoscale config is platform-specific+0.4
Ecosystem maturity (weight 8, our estimate) — established — our estimate from ecosystem age, docs and hiring pool+8
Scale-to-zero (weight 10, our estimate) — Serverless — you pay busy time only, idle costs nothing+10
Total (lines sum exactly)45.77
Possible range −64…+64 (sum of active rule weights) · fit % = (total + max) / (2 × max)(45.77 + 64) / 128 = 86%

How the $197.50+/mo estimate is calculated

H100 GPU-hours overage — 50hours over 0hours included × $3.95/hours$197.50
Note — $0.001097/sec = $3.95/hour (our ×3600 conversion); CPU ($0.0000131/core-sec) and memory ($0.00000222/GiB-sec) metered ON TOP — effective cost runs higher; A100-80GB $2.50/hr, T4 $0.59/hr$0
Monthly total (lines sum exactly)$197.50

What could change this verdict

Cost risk

Estimated now (StackSays estimate, not your final bill)$197.50/mo · H100 80GB serverless
Free-tier statusNo free tier — paid from the first unit
Biggest unpriced/uncertain variable$0.001097/sec = $3.95/hour (our ×3600 conversion); CPU ($0.0000131/core-sec) and memory ($0.00000222/GiB-sec) metered ON TOP — effective cost runs higher; A100-80GB $2.50/hr, T4 $0.59/hr

Official prices verified 2026-07-17; calculation is deterministic code over those prices.

At a glance

ToolFree tierPaid entryVerified
ModalNo free tierPay-as-you-go — $0.001097/sec = $3.95/hour (our ×3600 conversion); CPU ($0.0000131/core-sec) and memory ($0.00000222/GiB-sec) metered ON TOP — effective cost runs higher; A100-80GB $2.50/hr, T4 $0.59/hr2026-07-17
BeamNo free tierPay-as-you-go — $0.000986/sec = $3.55/hour (our ×3600 conversion); bills only while code runs, cold-start load time only; A100-80GB $2.25/hr, RTX 4090 $0.69/hr; CPU/RAM metered separately2026-07-17
RunPod ServerlessNo free tierPay-as-you-go — $4.55/hour serverless flex (their always-on pod H100 is $2.89/hr — the serverless premium is real); A100-80GB $2.72/hr, L40S-class $1.75/hr, cheapest 16GB SKUs $0.58/hr2026-07-17
ReplicateNo free tierPay-as-you-go — $0.001525/sec = $5.49/hour; A100-80GB $5.04/hr (≈2x Modal/Beam); T4 $0.81/hr; you're paying for the model-marketplace convenience, not raw compute value2026-07-17
BasetenNo free tierPay-as-you-go — $0.10833/min = $6.50/hour (our ×60 conversion); H100 MIG 40GB $3.75/hr, A100-80GB $4.00/hr, T4 $0.63/hr; per-minute increments round short bursts up harder than per-second rivals2026-07-17
FalNo free tierPay-as-you-go — $3.99/hour list (the 'as low as $1.89' column has no published qualifying conditions — treat list as the real on-demand rate); H200 $4.50/hr, B200 $6.25/hr; no T4/L4/A100 menu2026-07-17
LambdaNo free tierPay-as-you-go — $3.29/hour (H100 SXM $4.29); A100-40GB $1.99/hr; these are whole VMs — no scale-to-zero, so OUR HOURS = provisioned hours, not busy hours; prices exclude sales tax2026-07-17
Where did you land?anonymous, one click

Data sources

ToolSourceVerified
ModalModal pricing2026-07-17
BeamBeam pricing2026-07-17
RunPod ServerlessRunPod pricing2026-07-17
ReplicateReplicate pricing2026-07-17
BasetenBaseten pricing2026-07-17
FalFal pricing2026-07-17
LambdaLambda GPU Cloud pricing2026-07-17

A monitor re-checks these pages for changes; when a vendor moves a price, the date updates and every verdict recomputes.

Keep deciding

Compare head-to-head

RunPod Serverless vs Modal — for your app, not in general →Baseten vs Modal — for your app, not in general →Baseten vs Fal — for your app, not in general →RunPod Serverless vs Lambda — for your app, not in general →

Alternatives

Modal alternatives — ranked for your stack →Beam alternatives — ranked for your stack →RunPod Serverless alternatives — ranked for your stack →Baseten alternatives — ranked for your stack →Fal alternatives — ranked for your stack →Lambda alternatives — ranked for your stack →
AI Models & Infrastructure channel →