Hojo ASR V1
Hojo AI · released May 24, 2026 · HojoAI/Hojo-ASR-V1
Speech to textTranscribes a recording into words
- Type
- Open weightsApache License 2.0
- Languages
- 2
- Size
- 5.2B
Context measured in tokens
Our take
Written Aug 3, 2026Hojo ASR V1 is a 5.2-billion-parameter speech-to-text model you can download and run yourself, licensed under Apache 2.0. It transcribes clean audio with very few errors and processes an hour of audio in under a minute, but accuracy drops sharply on difficult recordings and no hosting providers currently offer it.
Pick this for self-hosted transcription with a permissive licence, especially clean read-aloud or financial calls where error stays under two percent. Use it for batch throughput — an hour of audio in 49 seconds. Skip it if you need meetings or accented speech, hosted inference, or languages beyond Chinese and English.
The case for it
- Extremely fast batch transcription: an hour of audio in 49 seconds on benchmark hardware.
- Strong on clean, structured audio: about one word in seventy wrong on read-aloud speech, and under two percent on financial calls.
- Permissive Apache 2.0 licence allows commercial use and redistribution.
- European-accented speech at 3.11% error — handled far better than its 8.15% on general accented speech.
The case against it
- Accuracy collapses on challenging real-world audio: meetings and accented speech see error rates more than six times higher than its clean-audio performance.
- No hosted inference available — zero offers in our catalogue, so self-hosting is mandatory.
- Only two languages listed, and every accuracy figure we hold is English; Chinese performance is unverified.
How good is it?
TranscriptionTurning speech into text5 of 5Open ASR WER · 2nd of 74
95.5%
Misses roughly one word in 22, averaged over nine English test sets.
73.8×60th of 62
an hour of audio in 49 seconds, on the board's own hardware. Your machine will differ.
2
Listed on the model card. The accuracy above is English only.
Percentage of words wrong on each set, lower better. Bars are scaled to this model's own worst case, not to the board.
Which languages ↓ ↑
Chinese · English
The figures above come from the Open ASR Leaderboard, an independent public test that runs every model on the same recordings. It is the only measurement of transcription quality we know of, so there are no other scores to show.
These tests check whether a model follows instructions — a precondition for all the work above, but not a measure of how well that work is done, which is why they get no rating.
Every published score for this model9 scoresEvery figure we hold, from 9 boards, with who ran it and a link to the source — including the boards no rating above is built on.
Can you run it yourself?
- Fits in memory
- weights load entirely on the card
- Spills to system RAM
- some weights offload; much slower
- Too large
- will not load even with offload
- est
- size is calculated; the verdict could change by 10%
Comfortable fit
GeForce RTX 4090 · 24 GB
Room to spare. 18.1 GB spare means a 10% error in the size would not change the answer.
GeForce RTX 5090 · 32 GB
Room to spare. 26.1 GB spare means a 10% error in the size would not change the answer.
Apple M2 (8-core GPU, 8GB unified) · 8 GB
Room to spare. 1.3 GB spare means a 10% error in the size would not change the answer.
Memory use by level
Against a 24 GB card.
All 3 sizes here are calculated, not measured. We hold no measured file for this model, so each size comes from the parameter count and every verdict above inherits that uncertainty.
Check against your own machine → · All 71 devices, with every size →
When we formed this view
Dates behind this page
Prices last checked 6h ago
What we do not know about this model yet
- We hold no measured file for it, so all 3 sizes on this page are calculated from the parameter count.
- Nothing we hold says whether an endpoint streams, so we do not show it either way.
Licence and identifiers
What the licence allowsApache License 2.0, what it allows commercially, and the identifiers you need to pull this model — its Hugging Face repo, our slug and a machine-readable card.
Licence
Apache License 2.0
Fully permissive: commercial use, redistribution, and derivatives allowed. Requires attribution and a copy of the license. Includes an express patent grant.
Identifiers
- Hugging Face
- HojoAI/Hojo-ASR-V1
- Modality record
- audio->text
- Catalogue slug
- hojoai-hojo-asr-v1