News
EveryLLM

Text

Reads and writes text. 471 models on EveryLLM so far.

Text is the base modality of every language model: it reads a prompt and writes a reply, token by token. A tokenizer first splits text into tokens, which are whole words, pieces of words or single characters; in English a token is roughly three-quarters of a word. Everything else a model does, from code to reasoning to calling tools, is ultimately expressed as text that it reads or writes.

Tokens and context

Most models use byte-pair encoding, which builds a vocabulary of frequent character sequences. Vocabularies have grown from about 50,000 tokens in GPT-2 to around 200,000 in GPT-4o and 256,000 in Gemma. The context window is how many tokens the model can consider at once: GPT-3 handled 2,048 in 2020, while Gemini 1.5 Pro offered one million in 2024. Languages in non-Latin scripts often need more tokens for the same meaning.

What text-only still means

Many models remain text-only, especially smaller ones and older families such as GPT-3, Llama 2 and BERT, and text still carries most business uses: drafting, summarising, pulling fields out of documents, translation and search. Text-only models are cheaper to train and serve, and easier to evaluate, since their inputs and outputs can be checked and compared directly.

Famous text models

Further reading

imgChitrarth 馃嚠馃嚦 IndiaChitrarth Multilingual vision-language model for ten Indian languages and English.Multimodal 路 Conditional licence Krutrim-2 馃嚠馃嚦 IndiaKrutrim-2 A 12-billion parameter multilingual language model focused on Indian languages and context.General 路 Open weights 路 2024 Sarvam 2B 馃嚠馃嚦 IndiaSarvam 2B Small language model for English and ten Indian languages.Small 路 Open weights 路 2024 ERNIE 3.0 Titan 馃嚚馃嚦 ChinaERNIE 3.0 Titan Large-scale Chinese knowledge-enhanced pre-trained language modelGeneral 路 Research only 路 2021 Fugaku-LLM 馃嚡馃嚨 JapanFugaku-LLM Japanese large language model trained on Fugaku supercomputerGeneral 路 Open weights 路 2024 fnTyphoon 馃嚬馃嚟 ThailandTyphoon Thai-focused large language model series by SCB 10XGeneral 路 Conditional licence 路 2024 Pragna-1B 馃嚠馃嚦 IndiaPragna-1B Small decoder model for Hindi, Bangla, Gujarati and EnglishSmall 路 Open weights Sailor 馃嚫馃嚞 SingaporeSailor Open multilingual language models for South-East AsiaGeneral 路 Open weights 路 2024 Sahabat-AI 馃嚠馃嚛 Indonesia and 1 moreSahabat-AI Indonesian language model family by GoTo and IndosatRegional 路 Conditional licence Aya Expanse WorldAya Expanse Open-weight multilingual language models in 8B and 32B sizes from Cohere For AIGeneral 路 Open weights 路 2024 Poro 馃嚝馃嚠 FinlandPoro Open 34B parameter model for Finnish, English and codeGeneral 路 Fully open 路 2024 Param-1 馃嚠馃嚦 IndiaParam-1 Bilingual and multilingual language models from India.General 路 Conditional licence AfriBERTa WorldAfriBERTa Multilingual encoder model for eleven African languagesEncoder 路 Open weights 路 2021 Airavata 馃嚠馃嚦 IndiaAiravata Instruction-tuned large language model for HindiRegional 路 Open weights 路 2024 GPT-SW3 馃嚫馃嚜 SwedenGPT-SW3 Generative language model for Nordic languagesGeneral 路 Research only 路 2023 Sarvam-Translate 馃嚠馃嚦 IndiaSarvam-Translate Document-level translation model for 22 Indian languages.Domain 路 Fully open Viking 馃嚝馃嚠 FinlandViking Open source language model for Nordic languages and codeGeneral 路 Fully open 路 2023 Meltemi 馃嚞馃嚪 GreeceMeltemi Open large language model for GreekGeneral 路 Open weights 路 2024 Aya 101 WorldAya 101 A 13-billion-parameter multilingual model that follows instructions in 101 languagesGeneral 路 Open weights 路 2024 HyperCLOVA 馃嚢馃嚪 South KoreaHyperCLOVA Korean language model with 82 billion parametersGeneral 路 Research only 路 2021 Nemotron-4-Mini-Hindi-4B-Instr 馃嚭馃嚫 United StatesNemotron-4-Mini-Hindi-4B-Instruct A small Hindi-focused chat model from NVIDIA, tuned for Indian-context answersSmall 路 Open weights 路 2024 ALLaM 馃嚫馃嚘 Saudi ArabiaALLaM Bilingual Arabic and English language model seriesGeneral 路 Open weights 路 2024 imgAya Vision WorldAya Vision Open-weights multilingual multimodal vision model from Cohere For AIMultimodal 路 Open weights

The weekly digest

Every Monday: the model of the day, new pages on EveryLLM, and the conferences and anniversaries of the week ahead.