Yanlan Model
Now in V3.0 — the new industry-leading bar for pre-publication Chinese correction, across text, video subtitles, audio transcription, and image OCR, built for production scale.
Yanlan, now in its third generation (V3.0), is a Chinese grammar-correction model trained with a multi-stage RL pipeline, built around the lived constraints of production deployment: false-alarm rates that don't waste editors' time, throughput that handles a daily news pipeline, and a cost profile that runs on a single mid-range GPU. The model handles plain text but also reads video subtitles, audio transcripts, and image OCR — so the corrections happen where the content lives.
Video subtitle correction
Hook Yanlan into the subtitle pipeline and it audits every line in stride — typos, mis-recognised characters, mismatched terms — at 99%+ accuracy and roughly 1–5 seconds per minute of video. Used by editorial desks shipping multilingual subtitles on a daily cadence.
Audio transcription correction
ASR transcripts arrive noisy. Yanlan reads them with the audio context in mind, fixes mis-heard homophones, normalises proper nouns, and respects dialectal forms instead of flattening them. 96%+ accuracy across major Mandarin variants.
Identity check on faces
Beyond text: Yanlan ships with a face-verification head trained for compliance review on broadcast and stream content. 98%+ recognition accuracy with a 20–30× speedup over the manual review pipeline.
Regulatory compliance check
Anchored to a corpus of 16,740 Chinese laws and regulations with daily auto-updates, Yanlan flags content that conflicts with current rules — useful for legal review on government, media, and education output.
See what changed in Yanlan V3.0.
All eight weighted headline metrics improve over V2.0, with 43 of 47 direct comparisons won. The full release brief separates the evidence, methodology, and deployment implications from the product overview.
Built to run in production
Accuracy is only half the job. A corrector that's accurate but slow never ships; one that's fast but expensive doesn't survive a finance review. Yanlan clears both bars on a single mid-range GPU, against the strongest commercial baselines we could reach.
How it works
Yanlan runs as a multimodal perception layer that normalises text, audio, image, and video subtitles into a common token stream, then routes through a classifier that selects between a central dictionary of standard Chinese, a custom dictionary tuned per deployment, and a correction model. A post-processing integration layer merges results before output.
Training is multi-stage RL on top of a Chinese-tuned base: supervised fine-tuning on human-verified corrections, then RL with editorial rewards that explicitly penalise false positives — that last step is what gets the false-alarm rate down to 0.5%.
Yanlan V3.0 is now fully available
Already serving news and publishing organizations, government communications, financial and legal compliance, and broadcast media. Enterprise and government teams: contact us to evaluate Yanlan on your own content.
Try Yanlan → Contact sales