Model library / Speech to text
Whisper large-v3: GPU requirements and hosting
A speech recognition checkpoint for transcription and speech translation. It serves an audio workflow rather than a text chat endpoint.
Sources reviewed 2026-09-30 · B3IQ engineering
Runtime profiles and requirements
The Xinference profile explicitly names large-v3. LocalAI uses a generic whisper-large artifact name, so confirm its resolved checkpoint before treating the profiles as equivalent.
xinference · audio-whisper-v3-xinference
- Catalog GPU threshold
- 6 GB
- Configured context
- 4,096 tokens
- Catalog system RAM
- 8 GB
- Runtime
- xinference
- Artifact identifier
whisper-large-v3- Precision
- Not pinned in catalog
- Configured concurrency
- 1 request(s)
- Profile inputs / outputs
- audio → text
- Deployed artifact revision
- Not verified
Catalog thresholds are configuration guidance, not measured peak memory. A positive GPU requirement is not recorded for every profile; unspecified is not zero. Context, concurrency and runtime overhead can increase memory use. Configured context is a profile setting, not the upstream maximum. Concurrency is configuration, not a load-test result.
localai · audio-whisper-large-localai
- Catalog GPU threshold
- Not specified
- Configured context
- 4,096 tokens
- Catalog system RAM
- 8 GB
- Runtime
- localai
- Artifact identifier
whisper-large- Precision
- Not pinned in catalog
- Configured concurrency
- 1 request(s)
- Profile inputs / outputs
- audio → text
- Deployed artifact revision
- Not verified
Catalog thresholds are configuration guidance, not measured peak memory. A positive GPU requirement is not recorded for every profile; unspecified is not zero. Context, concurrency and runtime overhead can increase memory use. Configured context is a profile setting, not the upstream maximum. Concurrency is configuration, not a load-test result.
Weight-only arithmetic and a planning estimate
The shared catalog does not record a parameter count for this profile. No weight-memory estimate is published.
This arithmetic does not describe the selected quantized artifact. It excludes a workload-specific KV-cache calculation and cannot guarantee fit at the configured context or concurrency. A mixture-of-experts model still stores its full weights.
See the assumptions →Plan the machine.
Memory-based machine candidates depend on the current visible store configurations. Confirm runtime compatibility, GPU count, interconnect and workload before purchase. Pricing and availability are shown on the machine page.
Explore machinesBefore deployment
Test the workload you need.
Measure word error rate and processing time on recordings with your accents, noise and language mix. Report audio duration and batching with any speed result.
These are catalog profiles and planning figures. No B3IQ performance measurement or live capacity is claimed. Confirm the artifact, runtime, workload and machine before deployment.
Evidence available
- Upstream identity / license
- Source reviewed
- Runtime settings
- Catalog configuration
- Memory fit
- Guidance; not a measured peak
- B3IQ performance
- No published benchmark
- Installation / live capacity
- Confirm for your machine
Source and access.
- Upstream checkpoint
- openai/whisper-large-v3 ↗
- Upstream license metadata
- Apache-2.0 ↗
- Access
- No access gate reported by the upstream repository at review.
- Reviewed source revision
06f233fe06e710322aca913c1bc4249a0d71fce1↗
The reviewed revision identifies the source used for this guide. It is not a claim that this revision is installed on a B3IQ machine. Review the publisher's current license and acceptable-use terms for your application.