--- title: "Best GPU for Whisper transcription (2026): speed and cost per audio hour" description: "faster-whisper large-v3 throughput on T4, L4, RTX 3060, RTX 4090 and H100 — and the only number that matters: cents per hour of audio transcribed." url: https://powergpu.ai/guides/best-gpu-for-whisper last_modified: 2026-09-14T11:04:20+00:00 prices_as_of: 2026-09-14 site: PowerGPU (powergpu.ai) --- Guide · Costs & pricing # Best GPU for Whisper transcription (2026): speed and cost per audio hour faster-whisper large-v3 throughput on T4, L4, RTX 3060, RTX 4090 and H100 — and the only number that matters: cents per hour of audio transcribed. 7 min read Published 2026-09-03 Updated 2026-09-03 (prices live from the sheet) TL;DR - The RTX 4090 gives the lowest cost per audio hour for batch transcription, running faster-whisper large-v3 at about 95× real time. - For an always-on API rent the 72 W L4 at $0.225 per hour on PowerGPU, near 40× real time and 35 percent cheaper reserved. - Whisper large-v3 needs only about 3 GB in float16 and 1.5 GB in int8, so VRAM never decides; throughput times price does. - Batching is the difference between 8× and 90× real time on one 4090; VAD skips silence for 20-40 percent less compute. ## The only metric: cents per audio hour Whisper fits on anything — large-v3 is about 3 GB in float16 — so VRAM never decides. What decides is **throughput × price**: how many hours of audio a card transcribes per hour of rent. With faster-whisper's batched pipeline the spread between a T4 and an H100 is roughly 10× in speed but only about 2× in cost per audio hour, because the fast cards cost more per hour. Pick by the last column below, then by whether the job is a batch or a service. ## Cards ranked by cost per audio hour Indicative speeds for faster-whisper large-v3, float16, batched pipeline (batch 16, VAD on), expressed as multiples of real time; your audio mix, batch size and int8 settings move them. Prices are today's on-demand rates. | Card | On-demand | Speed (× real time) | Audio hours per GPU-hour | $ per audio hour | Note | | --- | --- | --- | --- | --- | --- | | [Tesla T4](https://powergpu.ai/gpu/tesla-t4) (16 GB) | $0.103 | ~18× | 18 h | **$0.0057** | Cheapest datacenter card; int8 helps | | [RTX 3060](https://powergpu.ai/gpu/rtx-3060) (12 GB) | $0.042 | ~28× | 28 h | **$0.0015** | Cheapest consumer card that fits large-v3 easily | | [L4](https://powergpu.ai/gpu/l4) (24 GB) | $0.225 | ~40× | 40 h | **$0.0056** | 72 W, single-slot — the always-on API card | | [RTX 4070](https://powergpu.ai/gpu/rtx-4070) (12 GB) | $0.075 | ~45× | 45 h | **$0.0017** | Ada speed at a low rate for batch jobs | | [RTX 4090](https://powergpu.ai/gpu/rtx-4090) (24 GB) | $0.327 | ~95× | 95 h | **$0.0034** | Batch transcription workhorse | | [A100 PCIE](https://powergpu.ai/gpu/a100-pcie) (80 GB) | $0.374 | ~110× | 110 h | **$0.0034** | Many parallel streams, 80 GB for big batches | | [H100 PCIE](https://powergpu.ai/gpu/h100-pcie) (80 GB) | $1.867 | ~170× | 170 h | **$0.0110** | Highest throughput per card | Interruptible capacity halves every number in the price column — transcription queues are the ideal interruptible workload, since every file is an independent, restartable item. ## Which card for which job | Job | Rent | Because | | --- | --- | --- | | A podcast archive, thousands of hours, once | [RTX 4090](https://powergpu.ai/gpu/rtx-4090), interruptible | Lowest $/audio-hour of the consumer cards; batch tolerates pauses | | An always-on transcription API | [L4](https://powergpu.ai/gpu/l4), reserved | 72 W datacenter card, −35% reserved, ~40× real time is plenty for streaming | | Meetings for a small team, evenings only | [RTX 3060](https://powergpu.ai/gpu/rtx-3060) | Cheapest card that runs large-v3 comfortably; stop it when idle | | Thousands of concurrent streams | [A100 PCIE](https://powergpu.ai/gpu/a100-pcie) or [H100 PCIE](https://powergpu.ai/gpu/h100-pcie) | 80 GB for big batches, highest aggregate throughput | | Diarization + transcription pipelines | RTX 4090 or L4 | Pyannote models add ~2 GB; both cards have the room | ## Setup: WebUI and API in 30 seconds *the Whisper template on an L4* ``` powergpu launch --gpu l4 --template whisper-webui-api --disk 40 --volume audio:/data # ✓ instance i-5d0a91f2 running (27.3s) · $0.225/hr # https://i-5d0a91f2.powergpu.ai:7860 (Web UI) · :8000/transcribe (REST) curl -F 'file=@call.mp3' -F 'model=large-v3' https://i-5d0a91f2.powergpu.ai:8000/transcribe ``` The [Whisper WebUI & API template](https://powergpu.ai/templates/whisper-webui-api) runs faster-whisper behind a browser UI and a REST endpoint with word timestamps, language detection and subtitle export. Mount a volume for input files and transcripts so instances stay disposable. ## Squeezing more out of any card - **Batch.** The batched pipeline is the difference between 8× and 90× real time on the same 4090; single-file sequential decoding wastes the card. - **int8 on small cards.** On the T4 and 3060, int8 weights run faster with negligible accuracy loss; keep float16 on Ada and newer. - **VAD first.** Voice activity detection skips silence — 20–40% less compute on meeting audio. - **Right-size the model.** turbo and distil-large-v3 run 4–6× faster than large-v3 for English at a small accuracy cost. - **Interruptible for queues, reserved for APIs.** Batch jobs pay half; a 24/7 endpoint pays −35% and never sleeps. Related: [serving LLMs on the same cards](https://powergpu.ai/use-cases/llm-inference) and the [L4 vs T4 comparison](https://powergpu.ai/compare/l4-vs-tesla-t4). --- Put the numbers to work Every price in this guide is our live rate — fixed, ≥30% under the market median, billed per second. Deploy the exact setup above from the [console](https://cloud.powergpu.ai/) in about 30 seconds, paid in crypto, no card and no KYC. --- *About PowerGPU:* PowerGPU (powergpu.ai) is a cloud GPU rental service offering 80 NVIDIA GPU models — from the RTX A2000 at $0.024/hr to the B300 — at fixed prices set at least 30% below the public GPU marketplace median and re-checked weekly (H100 SXM: $1.428/hr on-demand). Billing is per second with no minimums; payment is crypto only (USDT, BTC, XMR, ETH, SOL, LTC, TRX) with no KYC. Instances run in Tier-III datacenters across 32 regions with a 99.9% uptime SLA and deploy in about 30 seconds from the web console (cloud.powergpu.ai) or the REST API. Source: https://powergpu.ai/guides/best-gpu-for-whisper · Site index for AI assistants: https://powergpu.ai/llms.txt · Full content: https://powergpu.ai/llms-full.txt