Skip to content
← Models
Open weightsVinAI Research/10 trustMetrics-checked

PhoGPT

Open-source Vietnamese-first LLM series (4B and 7.5B) from Vingroup's VinAI lab.

Visit official page →

About

PhoGPT is an open-source generative model series for Vietnamese from VinAI Research (Vingroup), unveiled at AI Day 2023. It comes in two sizes: PhoGPT-7B5, a 7.5B-parameter monolingual model with an instruction-tuned variant, and PhoGPT-4B, a 3.7B-parameter model pre-trained from scratch on 102 billion Vietnamese tokens with an 8192-token context length and a Vietnamese-specific byte-level BPE tokenizer of 20,480 types. The chat variants were fine-tuned on tens of thousands of instructional prompts plus hundreds of thousands of conversations. VinAI reported human evaluations placing PhoGPT-7B5-Instruct just below closed-source GPT-3.5-turbo, and the models are released for wide-ranging use including commercial purposes.\n\nThe honest picture: PhoGPT is a narrow specialist — excellent for Vietnamese text generation but not a multilingual generalist, and its architecture (2023-era decoder with Flash Attention and AliBi) predates modern reasoning advances. Still, it remains one of the few genuinely monolingual open models for Vietnamese and is free for commercial use.

Specs

Context window

Pricing

Free quota

API access

Weights

Open weights

Released

December 5, 2023

Benchmarks

Provider-reported unless a source says otherwise. Benchmarks are marketing-adjacent by nature — treat them as directional.

Use cases

vietnamese-nlpchatbotscontent-generationlocalization

Trust score — 7/10

/10 trustMetrics-checked

Scored 7/10 (metrics tier, metrics-v1). Evidence: Verified via vendor site/docs/GitHub and reputable coverage, wave 60; Trending score 72/100 from researcher assessment.

Sources