Hy-SimulMT is a real-time simultaneous translation model developed by Tencent Hunyuan, supporting bidirectional speech-to-speech translation between Chinese and English. It continuously receives incoming audio and translates as the speaker talks, producing both translated text and speech.
Alternatives to Hy-SimulMT-30B-A3B
72 open-weight models that do the same job as Hy-SimulMT-30B-A3B (30.4B translation), ranked by how close they are in task and size, then by popularity and how recent they are. Licence and VRAM are shown for each.
This is a COMET quality estimation model: It receives a source sentence and the respective translation and returns a score that reflects the quality of the translation.
?BCC BY-NC-SA 4.04K 128Feb 2023- Same category
Hunyuan Translation Model Version 1.5 includes a 1.8B translation model, HY-MT1.5-1.8B, and a 7B translation model, HY-MT1.5-7B. Both models focus on supporting mutual translation across 33 languages and incorporating 5 ethnic and dialect variations.
2B256K ctx Unknown licence4.8K 1.2KDec 2025- Same category
- Smaller (2B vs 30.4B)
North Small Translate is an open weights research release of a sparse Mixture-of-Experts model with 25 billion active parameters and 218 billion total parameters, specialized for high-quality machine translation across 50 languages.
218BCC BY-NC 4.032 67Aug 2026- Same category
- Larger (218B vs 30.4B)
This is the model card of NLLB-200's distilled 600M variant.
600M1K ctx CC BY-NC 4.0732.9K 1KJul 2022- Same category
- Smaller (600M vs 30.4B)
This is the model card of IndicTrans2 En-Indic Distilled 200M variant.
275MMIT310.2K 100Sept 2023- Same category
- Smaller (275M vs 30.4B)
- Commercial use OK
opus-mt-fr-en is an open-weight translation model with 75.2M parameters from Helsinki-NLP, published on Hugging Face.
75M512 ctx Apache 2.0333.5K 61Mar 2022- Same category
- Smaller (75M vs 30.4B)
- Commercial use OK
This is the model card of NLLB-200's distilled 1.3B variant.
1.3B1K ctx CC BY-NC 4.0298.7K 195Jul 2022- Same category
- Smaller (1.3B vs 30.4B)
TowerInstruct-7B is a language model that results from fine-tuning TowerBase on the TowerBlocks supervised fine-tuning dataset. TowerInstruct-7B-v0.1 is the first model in the series.
6.7B4K ctx CC BY-NC 4.0115 68Jan 2024- Same category
- Smaller (6.7B vs 30.4B)
Based on the LFM2-350M model, this checkpoint has been fine-tuned for near real-time bi-directional Japanese/English translation of short-to-medium inputs.
354M125K ctx lfm1.01K 99Sept 2025- Same category
- Smaller (354M vs 30.4B)
This is the model card of IndicTrans2 En-Indic 1.1B variant.
1.1BMIT49K 95Sept 2023- Same category
- Smaller (1.1B vs 30.4B)
- Commercial use OK
Hibiki is a model for streaming speech translation (also known as simultaneous translation). Unlike offline translation—where one waits for the end of the source utterance to start translating--- Hibiki adapts its flow to accumulate just enough context to produce a correct translation in real-time…
2BCC BY 4.01.5K 64Feb 2025- Same category
- Smaller (2B vs 30.4B)
- Commercial use OK
New multilingual machine translation model (MiLMMT-46) is now available. Please check the link for detailed information.
3.2B8K ctx Gemma Terms of Use103 113Nov 2024- Same category
- Smaller (3.2B vs 30.4B)
Data Freshness: The pre-training data has a cutoff date of June 25, 2025.
63.2BA3BNVIDIA Open Model License726 148Apr 2026- Related category
- Larger (63.2B vs 30.4B)
- Commercial use OK
Tower+ 9B is build on top of Gemma 2 9B. The model goes through the Continuous Pretraining (CPT), Instruction Tuning (IT), Weighted Preference Optimization (WPO). During all stages we include parallel and multilingual data (covering 22 languages).
9.2B8K ctx CC BY-NC-SA 4.016.9K 43Jun 2025- Related category
- Smaller (9.2B vs 30.4B)
Voxtral Mini 4B Realtime 2602 is a multilingual, realtime speech-transcription model and among the first open-source solutions to achieve accuracy comparable to offline systems with a delay of <500ms.
4.4BApache 2.01.5M 1.1KJan 2026- Related category
- Smaller (4.4B vs 30.4B)
- Commercial use OK
- Widely used
The Qwen3-ASR family includes Qwen3-ASR-1.7B and Qwen3-ASR-0.6B, which support language identification and ASR for 52 languages and dialects. Both leverage large-scale speech training data and the strong audio understanding capability of their foundation model, Qwen3-Omni.
782MApache 2.0535K 99Jun 2026- Related category
- Smaller (782M vs 30.4B)
- Commercial use OK
The Qwen3-ASR family includes Qwen3-ASR-1.7B and Qwen3-ASR-0.6B, which support language identification and ASR for 52 languages and dialects. Both leverage large-scale speech training data and the strong audio understanding capability of their foundation model, Qwen3-Omni.
2.3BApache 2.01.6M 1.3KJan 2026- Related category
- Smaller (2.3B vs 30.4B)
- Commercial use OK
- Widely used
The Qwen3-ASR family includes Qwen3-ASR-1.7B and Qwen3-ASR-0.6B, which support language identification and ASR for 52 languages and dialects. Both leverage large-scale speech training data and the strong audio understanding capability of their foundation model, Qwen3-Omni.
2BApache 2.0347.7K 170Jun 2026- Related category
- Smaller (2B vs 30.4B)
- Commercial use OK
Voxtral Mini 4B Realtime Arabic is a streaming speech-to-text model for Arabic dialects and Modern Standard Arabic. It is designed to handle challenging real-world audio and code-switching, where speakers alternate between languages within the same conversation.
4.4BUnknown licence2.2K 56Oct 2026- Related category
- Smaller (4.4B vs 30.4B)
The Qwen3-ASR family includes Qwen3-ASR-1.7B and Qwen3-ASR-0.6B, which support language identification and ASR for 52 languages and dialects. Both leverage large-scale speech training data and the strong audio understanding capability of their foundation model, Qwen3-Omni.
938MApache 2.0232.4K 400Jan 2026- Related category
- Smaller (938M vs 30.4B)
- Commercial use OK
Whisper is a state-of-the-art model for automatic speech recognition (ASR) and speech translation, proposed in the paper Robust Speech Recognition via Large-Scale Weak Supervision by Alec Radford et al. from OpenAI.
809MMIT4.7M 3.6KOct 2024- Related category
- Smaller (809M vs 30.4B)
- Commercial use OK
- Widely used
The most accurate verbatim speech recognition you can run in production: controllable, multilingual, and timed to the word.
1.5Bnyra-health-non-commercial-research24.6K 123Jul 2026- Related category
- Smaller (1.5B vs 30.4B)
Canary-1b-v2 is a powerful 1-billion parameter model built for high-quality speech transcription and translation across 25 European languages.
979MCC BY 4.0228.6K 451Aug 2025- Related category
- Smaller (979M vs 30.4B)
- Commercial use OK
Cohere Labs Command A Translate is an open weights research release of a 111 billion parameter model that achieves state-of-the-art performance on translation quality.
111BCC BY-NC 4.014 79Aug 2025- Related category
- Larger (111B vs 30.4B)