Software:List of large language models

From HandWiki
Short description: none


This is a dynamic list and may never be able to satisfy particular standards for completeness. You can help by expanding it with reliably sourced entries.

A large language model (LLM) is a type of machine learning model designed for natural language processing tasks such as language generation. LLMs are language models with many parameters, and are trained with self-supervised learning on a vast amount of text.

This page lists notable large language models.

List

For the training cost column, 1 petaFLOP-day = 1 petaFLOP/sec × 1 day = 8.64E19 FLOP. Also, only the largest model's cost is written.

Template:Sort-under

Name Release date[lower-alpha 1] Developer Number of parameters (billion) [lower-alpha 2] Corpus size Training cost (petaFLOP-day) License[lower-alpha 3] Notes
Attention Is All You Need 2017-06 Vaswani et al at Google 0.213 36 million English-French sentence pairs 0.09[1] Unreleased Trained for 0.3M steps on 8 NVIDIA P100 GPUs. Training and evaluation code released under Apache 2.0 license.[2]
GPT-1 2018-06 OpenAI 0.117 0.117


Unknown 1[3] MIT[4] First GPT model, decoder-only transformer. Trained for 30 days on 8 P600 GPUs.
BERT 2018-10 Google 0.340 0.340

[5] || 3300000000 3.3 billion 

words[5]
9 9

[6]|| style="background:#9F9;vertical-align:middle;text-align:center;" class="table-yes"|Apache 2.0[7]

T5 2019-10 Google 11 11

[8]

34 billion tokens[8] Apache 2.0[9] Base model for many Google projects, such as Imagen.[10]
XLNet 2019-06 Google 0.340 0.340

[11]|| 3300000000 33 

billion words
330 Apache 2.0[12] An alternative to BERT; designed as encoder-only. Trained on 512 TPU v3 chips for 5.5 days.[13]
GPT-2 2019-02 OpenAI 1.5 1.5

[14] || 40GB[15] (~10000000000 10 billion 

tokens)[16]
28[17] MIT[18] Trained on 32 TPUv3 chips for 1 week.[17]
GPT-3 2020-05 OpenAI 175 175

[19] || 300000000000 300 billion 

tokens[16]
3640[20] Proprietary A fine-tuned variant of GPT-3, termed GPT-3.5, was made available to the public through a web interface called ChatGPT in 2022.[21]
GPT-Neo 2021-03 EleutherAI 2.7 2.7

[22]

825 GiB[23] Unknown MIT[24] The first of a series of free GPT-3 alternatives released by EleutherAI. GPT-Neo outperformed an equivalent-size GPT-3 model on some benchmarks, but was significantly worse than the largest GPT-3.[24]
GPT-J 2021-06 EleutherAI 6 6

[25] || 825 GiB[23]

200[26] Apache 2.0 GPT-3-style language model
Megatron-Turing NLG 2021-10[27] Microsoft and Nvidia 530 530

[28]

338600000000 338.6 billion 
tokens[28]
38000[29] Unreleased Trained for 3 months on over 2000 A100 GPUs on the NVIDIA Selene Supercomputer, for over 3 million GPU-hours[29]
Ernie 3.0 Titan 2021-12 Baidu 260 260

[30]

4TB Unknown Proprietary Chinese-language LLM. Ernie Bot is based on this model.
Claude[31] 2021-12 Anthropic 52 52

[32]

400000000000 400 billion 
tokens[32]
Unknown Proprietary Fine-tuned for desirable behavior in conversations.[33]
GLaM (Generalist Language Model) 2021-12 Google 1200 1200

[34] || 1600000000000 1.6 trillion 

tokens[34]
5600[34] Proprietary Sparse mixture of experts model, making it more expensive to train but cheaper to run inference compared to GPT-3.
Gopher 2021-12 DeepMind 280 280

[35] || 300000000000 300 billion 

tokens[36]
5833[37] Proprietary Later developed into the Chinchilla model.
LaMDA (Language Models for Dialog Applications) 2022-01 Google 137 137

[38] || 1.56T words,[38] 168000000000 168 billion 

tokens[36]
4110[39] Proprietary Specialized for response generation in conversations.
GPT-NeoX 2022-02 EleutherAI 20 20

[40] || 825 GiB[23]

740[26] Apache 2.0 based on the Megatron architecture
Chinchilla 2022-03 DeepMind 70 70

[41] || 1400000000000 1.4 trillion 

tokens[41][36]
6805[37] Proprietary Reduced-parameter model trained on more data. Used in the Sparrow bot. Often cited for its neural scaling law.
PaLM (Pathways Language Model) 2022-04 Google 540 540

[42] || 768000000000 768 billion 

tokens[41]
29250 29,250

[37]|| style="background: #ddf; vertical-align: middle; text-align: center; " class="table-proprietary"|Proprietary

Trained for ~60 days on ~6000 TPU v4 chips.[37]
OPT (Open Pretrained Transformer) 2022-05 Meta 175 175

[43] || 180000000000 180 billion 

tokens[44]
310[26] Non-commercial research[lower-alpha 4] GPT-3 architecture with some adaptations from Megatron. Uniquely, the training logbook written by the team was published.[45]
YaLM 100B 2022-06 Yandex 100 100

[46]

1.7TB[46] Unknown Apache 2.0 English-Russian model based on Microsoft's Megatron-LM
Minerva 2022-06 Google 540 540

[47]

38.5B tokens from webpages filtered for mathematical content and from papers submitted to the arXiv preprint server[47] Unknown Proprietary For solving "mathematical and scientific questions using step-by-step reasoning".[48] Initialized from PaLM models, then finetuned on mathematical and scientific data.
BLOOM 2022-07 Large collaboration led by Hugging Face 175 175

[49]

350000000000 350 billion 
tokens (1.6TB)[50]
Unknown Responsible AI Essentially GPT-3 but trained on a multi-lingual corpus (30% English excluding programming languages)
Galactica 2022-11 Meta 120 120 350000000000 106 billion 
tokens[51]
Unknown CC-BY-NC-4.0 Trained on scientific text and modalities.
AlexaTM (Teacher Models) 2022-11 Amazon 20 20

[52] || 1300000000000 1.3 trillion

[53]

Unknown Proprietary[54] Bidirectional sequence-to-sequence architecture
Llama 2023-02 Meta AI 65 65

[55] || 1400000000000 1.4 trillion

[55]

6300[56] Non-commercial research[lower-alpha 5] Corpus has 20 languages. "Overtrained" (compared to Chinchilla scaling law) for better performance with fewer parameters.[55]
GPT-4 2023-03 OpenAI Unknown[lower-alpha 6]
(According to rumors: 1760)[58] 		Unknown Unknown,
estimated 230,000 		Proprietary Available for all ChatGPT users now and used in several products.
Chameleon 2024-06 Meta AI 34 34

[59]

4400000000000 4.4 trillion


Unknown Non-commercial research[60]
Cerebras-GPT 2023-03 Cerebras 13 13

[61]

270[26] Apache 2.0 Trained with Chinchilla formula.
Falcon 2023-03 Technology Innovation Institute 40 40

[62] || 1 trillion tokens, from RefinedWeb (filtered web text corpus)[63] plus some "curated corpora".[64]

2800[56] Apache 2.0[65]
BloombergGPT 2023-03 Bloomberg L.P. 50 50 363 billion token dataset based on Bloomberg's data sources, plus 345 billion tokens from general purpose datasets[66] Unknown Unreleased Trained on financial data from proprietary sources, for financial tasks
PanGu-Σ 2023-03 Huawei 1085 1085


329 billion tokens[67] Unknown Proprietary
OpenAssistant[68] 2023-03 LAION 17 17


1.5 trillion tokens Unknown Apache 2.0 Trained on crowdsourced open data
Jurassic-2[69] 2023-03 AI21 Labs Unknown Unknown Unknown Proprietary Multilingual[70]
PaLM 2 (Pathways Language Model 2) 2023-05 Google 340 340

[71] || 3600000000000 3.6 trillion 

tokens[71]
85000 85,000

[56]|| style="background: #ddf; vertical-align: middle; text-align: center; " class="table-proprietary"|Proprietary

Was used in Bard chatbot.[72]
Llama 2 2023-07 Meta AI 70 70

[73] || 2000000000000 2 trillion 

tokens[73]
21000 21,000 style="background: #FFB; color: black; vertical-align: middle; text-align: center; " class="table-partial" | Llama 2 license 1.7 million A100-hours.[74]
Claude 2 2023-07 Anthropic Unknown Unknown Unknown Proprietary Used in Claude chatbot.[75]
Granite 13b 2023-07 IBM Unknown Unknown Unknown Proprietary Used in IBM Watsonx.[76]
Mistral 7B 2023-09 Mistral AI 7.3 7.3

[77]

Unknown Unknown Apache 2.0
Claude 2.1 2023-11 Anthropic Unknown Unknown Unknown Proprietary Used in Claude chatbot. Has a context window of 200,000 tokens, or ~500 pages.[78]
Grok 1[79] 2023-11 xAI 314 Unknown Unknown Apache 2.0 Used in Grok chatbot. Grok 1 has a context length of 8,192 tokens and has access to X (Twitter).[80]
Gemini 1.0 2023-12 Google DeepMind Unknown Unknown Unknown Proprietary Multimodal model, comes in three sizes. Used in the chatbot of the same name.[81]
Mixtral 8x7B 2023-12 Mistral AI 46.7 Unknown Unknown Apache 2.0 Outperforms GPT-3.5 and Llama 2 70B on many benchmarks.[82] Mixture of experts model, with 12.9 billion parameters activated per token.[83]
Mixtral 8x22B 2024-04 Mistral AI 141 Unknown Unknown Apache 2.0 [84]
DeepSeek-LLM Template:DTS DeepSeek 67 2T tokens[85]Template:Pg 12000 12,000


DeepSeek License Trained on English and Chinese text. 1e24 FLOPs for 67B. 1e23 FLOPs for 7B[85]Template:Pg
Phi-2 2023-12 Microsoft 2.7 1.4T tokens 419[86] MIT Trained on real and synthetic "textbook-quality" data, for 14 days on 96 A100 GPUs.[86]
Gemini 1.5 2024-02 Google DeepMind Unknown Unknown Unknown Proprietary Multimodal model, based on a Mixture-of-Experts (MoE) architecture. Context window above 1 million tokens.[87]
Gemini Ultra 2024-02 Google DeepMind Unknown Unknown Unknown Proprietary
Gemma 2024-02 Google DeepMind 7 6T tokens Unknown Gemma Terms of Use[88]
Claude 3 2024-03 Anthropic Unknown Unknown Unknown Proprietary Includes three models, Haiku, Sonnet, and Opus.[89]
DBRX 2024-03 Databricks and Mosaic ML 136 136


12T tokens Unknown Databricks Open Model License[90][91] Training cost 10 million USD
Fugaku-LLM 2024-05 Fujitsu, Tokyo Institute of Technology, etc. 13 13


380B tokens Unknown Fugaku-LLM Terms of Use[92] The largest model ever trained on CPU-only, on the Fugaku[93]
Phi-3 2024-04 Microsoft 14[94] 4.8T tokens Unknown MIT Microsoft markets them as "small language model".[95]
Granite Code Models 2024-05 IBM Unknown Unknown Unknown Apache 2.0
Qwen2 2024-06 Alibaba Cloud 72[96] 3T tokens Unknown Qwen License Multiple sizes, the smallest being 0.5B.
DeepSeek-V2 Template:DTS DeepSeek 236 8.1T tokens 28000 28,000


DeepSeek License 1.4M hours on H800.[97]
Nemotron-4 2024-06 Nvidia 340 340


9T tokens 200000 200,000


NVIDIA Open Model License[98][99] Trained for 1 epoch. Trained on 6144 H100 GPUs between December 2023 and May 2024.[100][101]
Claude 3.5 2024-06 Anthropic Unknown Unknown Unknown Proprietary Initially, only one model, Sonnet, was released.[102] In October 2024, Sonnet 3.5 was upgraded, and Haiku 3.5 became available.[103]
Llama 3.1 2024-07 Meta AI 405 15.6T tokens 440000 440,000


Llama 3 license 405B version took 31 million hours on H100-80GB, at 3.8E25 FLOPs.[104][105]
OpenAI o1 2024-09-12 OpenAI Unknown Unknown Unknown Proprietary Reasoning model.[106]
Mistral Large 2024-11 Mistral AI 123 Unknown Unknown Mistral Research License Upgraded over time. The latest version is 24.11.[107]
Pixtral 2024-11 Mistral AI 123 Unknown Unknown Mistral Research License Multimodal. There is also a 12B version which is under Apache 2 license.[107]
DeepSeek-V3 2024-12 DeepSeek 671 14.8T tokens 56000 56,000


MIT 2.788M hours on H800 GPUs.[108] Originally released under the DeepSeek License, then re-released under the MIT License as "DeepSeek-V3-0324" in March 2025.[109]
Amazon Nova 2024-12 Amazon Unknown Unknown Unknown Proprietary Includes three models, Nova Micro, Nova Lite, and Nova Pro[110]
DeepSeek-R1 2025-01 DeepSeek 671 Not applicable Unknown MIT No pretraining. Reinforcement-learned upon V3-Base.[111][112]
Qwen2.5 2025-01 Alibaba 72 18T tokens Unknown Qwen License 7 dense models, with parameter count from 0.5B to 72B. They also released 2 MoE variants.[113]
MiniMax-Text-01 2025-01 Minimax 456 4.7T tokens[114] Unknown Minimax Model license [115][114]
Gemini 2.0 2025-02 Google DeepMind Unknown Unknown Unknown Proprietary Three models released: Flash, Flash-Lite and Pro[116][117][118]
Claude 3.7 2025-02-24 Anthropic Unknown Unknown Unknown Proprietary One model, Sonnet 3.7.[119]
GPT-4.5 2025-02-27 OpenAI Unknown Unknown Unknown Proprietary Largest non-reasoning model.[120]
Grok 3 2025-02 xAI Unknown Unknown Unknown,
estimated 5,800,000 		Proprietary Training cost claimed "10x the compute of previous state-of-the-art models".[121]
Gemini 2.5 2025-03-25 Google DeepMind Unknown Unknown Unknown Proprietary Three models released: Flash, Flash-Lite and Pro[122]
Llama 4 2025-04-05 Meta AI 400 400


40000000000000 40T tokens


Unknown Llama 4 license [123][124]
OpenAI o3 and o4-mini 2025-04-16 OpenAI Unknown Unknown Unknown Proprietary Reasoning models.[125]
Qwen3 2025-04 Alibaba Cloud 235 36000000000000 36T tokens


Unknown Apache 2.0 Multiple sizes, the smallest being 0.6B.[126]
Claude 4 2025-05-22 Anthropic Unknown Unknown Unknown Proprietary Includes two models, Sonnet and Opus.[127]
Grok 4 2025-07-09 xAI Unknown Unknown Unknown Proprietary
GLM-4.5 2025-07-29 Zhipu AI 355 22T tokens Unknown MIT Released in 335B and 106B sizes.[128] Corpus size was calculated by combining the 15 trillion tokens and the 7 trillion tokens pre-training mix.[129]
GPT-OSS 2025-08-05 OpenAI 117 Unknown Unknown Apache 2.0 Released in 20B and 120B sizes.[130]
Claude 4.1 2025-08-05 Anthropic Unknown Unknown Unknown Proprietary Includes one model, Opus.[131]
GPT-5 2025-08-07 OpenAI Unknown Unknown Unknown Proprietary Includes three models, GPT-5, GPT-5 mini, and GPT-5 nano. GPT-5 is available in ChatGPT and API. It includes thinking abilities. [132][133]
DeepSeek-V3.1 August 21, 2025 DeepSeek 671 15.639T MIT Training size: 14.8T tokens, of DeepSeek V3 plus 839B tokens from the extension phases (630B + 209B)[134]It is a hybrid model that can switch between thinking and non-thinking modes.[135]
Claude 4.5 2025-09-29 Anthropic Unknown Unknown Unknown Proprietary Only one variant is available, Sonnet.[136]
DeepSeek-V3.2-Exp 2025-09-29 DeepSeek 685 MIT This experimental model built upon v3.1-Terminus uses a custom efficient mechanism tagged DeepSeek Sparse Attention (DSA).[137][138][139]
GLM-4.6 2025-09-30 Zhipu AI 357 Apache 2.0 [140][141][142]

Timeline

Timeline of major LLM releases (2024–present)
<timeline>

DateFormat=mm/dd/yyyy Define $start = 01/01/2024 Define $now = 02/08/2026 Define $later = 08/08/2026 Define $dayunknown = 15 Period = from:$start till:$later ImageSize= width:1250 height:auto barincrement:20 TimeAxis = orientation:horizontal PlotArea = right:5 left:5 bottom:130 top:10

Colors = 

 id:bg        value:white
 id:lightline value:rgb(0.9,0.9,0.9)
 id:lighttext value:rgb(0.5,0.5,0.5)
 id:openai    value:rgb(0.60,0.45,0.85)   Legend:OpenAI
 id:google    value:rgb(0.26,0.52,0.96)   Legend:Google_(Gemini)
 id:anthropic value:rgb(0.95,0.55,0.20)   Legend:Anthropic_(Claude)
 id:meta      value:rgb(0.20,0.70,0.40)   Legend:Meta_(Llama)
 id:mistral   value:rgb(0.20,0.70,0.70)   Legend:Mistral
 id:xai       value:rgb(0.50,0.50,0.50)   Legend:xAI_(Grok)
 id:cohere    value:rgb(0.95,0.30,0.60)   Legend:Cohere

BackgroundColors = canvas:bg ScaleMajor = gridcolor:lighttext unit:month increment:1 start:$start ScaleMinor = gridcolor:lightline unit:day increment:14 start:$start Legend = orientation:vertical position:bottom columns:1

BarData = 

 barset:OpenAI
 barset:Google
 barset:Anthropic
 barset:Meta
 barset:Mistral
 barset:xAI
 barset:Cohere

PlotData = 

 width:15 textcolor:black shift:(5,-5) anchor:from fontsize:s

 barset:OpenAI
   at:05/13/2024 text:"GPT-4o"
   color:openai from:05/13/2024 till:08/07/2025 text:""
   at:08/07/2025 text:"GPT-5"
   color:openai from:08/07/2025 till:$now text:""

 barset:Google
   at:02/15/2024 text:"Gemini 1.5"
   color:google from:02/15/2024 till:12/11/2024 text:""
   at:12/11/2024 text:"Gemini 2.0"
   color:google from:12/11/2024 till:$now text:""

 barset:Anthropic
   at:03/04/2024 text:"Claude 3 (family)"
   color:anthropic from:03/04/2024 till:06/20/2024 text:""
   at:06/20/2024 text:"Claude 3.5 Sonnet"
   color:anthropic from:06/20/2024 till:$now text:""

 barset:Meta
   at:04/18/2024 text:"Llama 3"
   color:meta from:04/18/2024 till:07/23/2024 text:""
   at:07/23/2024 text:"Llama 3.1"
   color:meta from:07/23/2024 till:$now text:""

 barset:Mistral
   at:02/26/2024 text:"Mistral Large"
   color:mistral from:02/26/2024 till:$now text:""

 barset:xAI
   at:03/28/2024 text:"Grok-1.5"
   color:xai from:03/28/2024 till:12/14/2024 text:""
   at:12/14/2024 text:"Grok-2"
   color:xai from:12/14/2024 till:$now text:""

 barset:Cohere
   at:03/24/2024 text:"Command R"
   color:cohere from:03/24/2024 till:08/15/2024 text:""
   at:08/15/2024 text:"Command R+ (08-2024)"
   color:cohere from:08/15/2024 till:$now text:""

</timeline>


See also

Notes

  1. ↑ This is the date that documentation describing the model's architecture was first released.
  2. ↑ In many cases, researchers release or report on multiple versions of a model having different sizes. In these cases, the size of the largest model is listed here.
  3. ↑ This is the license of the pre-trained model weights. In almost all cases the training code itself is open-source or can be easily replicated.
  4. ↑ The smaller models including 66B are publicly available, while the 175B model is available on request.
  5. ↑ Facebook's license and distribution scheme restricted access to approved researchers, but the model weights were leaked and became widely available.
  6. ↑ As stated in Technical report: "Given both the competitive landscape and the safety implications of large-scale models like GPT-4, this report contains no further details about the architecture (including model size), hardware, training compute, dataset construction, training method ..."[57]

References

  1. ↑ "AI and compute" (in en-US). 2022-06-09. https://openai.com/index/ai-and-compute/. 
  2. ↑ "Apache License" (in en). TensorFlow. https://github.com/tensorflow/tensor2tensor/blob/3d9c62f2aca9492db5c22676416974005b9dcbae/LICENSE. 
  3. ↑ "Improving language understanding with unsupervised learning". June 11, 2018. https://openai.com/research/language-unsupervised. 
  4. ↑ "finetune-transformer-lm". GitHub. https://github.com/openai/finetune-transformer-lm. 
  5. ↑ 5.0 5.1 Devlin, Jacob; Chang, Ming-Wei; Lee, Kenton; Toutanova, Kristina (11 October 2018). "BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding". arXiv:1810.04805v2 [cs.CL].
  6. ↑ Prickett, Nicole Hemsoth (2021-08-24). "Cerebras Shifts Architecture To Meet Massive AI/ML Models". https://www.nextplatform.com/2021/08/24/cerebras-shifts-architecture-to-meet-massive-ai-ml-models/. 
  7. ↑ "BERT". March 13, 2023. https://github.com/google-research/bert. 
  8. ↑ 8.0 8.1 Raffel, Colin; Shazeer, Noam; Roberts, Adam; Lee, Katherine; Narang, Sharan; Matena, Michael; Zhou, Yanqi; Li, Wei et al. (2020). "Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer". Journal of Machine Learning Research 21 (140): 1–67. ISSN 1533-7928. http://jmlr.org/papers/v21/20-074.html. 
  9. ↑ google-research/text-to-text-transfer-transformer, Google Research, 2024-04-02, https://github.com/google-research/text-to-text-transfer-transformer, retrieved 2024-04-04 
  10. ↑ "Imagen: Text-to-Image Diffusion Models". https://imagen.research.google/. 
  11. ↑ "Pretrained models — transformers 2.0.0 documentation". https://huggingface.co/transformers/v2.0.0/pretrained_models.html. 
  12. ↑ "xlnet". GitHub. https://github.com/zihangdai/xlnet/. 
  13. ↑ Yang, Zhilin; Dai, Zihang; Yang, Yiming; Carbonell, Jaime; Salakhutdinov, Ruslan; Le, Quoc V. (2 January 2020). "XLNet: Generalized Autoregressive Pretraining for Language Understanding". arXiv:1906.08237 [cs.CL].
  14. ↑ "GPT-2: 1.5B Release" (in en). 2019-11-05. https://openai.com/blog/gpt-2-1-5b-release/. 
  15. ↑ "Better language models and their implications". https://openai.com/research/better-language-models. 
  16. ↑ 16.0 16.1 "OpenAI's GPT-3 Language Model: A Technical Overview". 3 June 2020. https://lambdalabs.com/blog/demystifying-gpt-3. 
  17. ↑ 17.0 17.1 "openai-community/gpt2-xl · Hugging Face". https://huggingface.co/openai-community/gpt2-xl. 
  18. ↑ "gpt-2". GitHub. https://github.com/openai/gpt-2. 
  19. ↑ Wiggers, Kyle (28 April 2022). "The emerging types of language models and why they matter". TechCrunch. https://techcrunch.com/2022/04/28/the-emerging-types-of-language-models-and-why-they-matter/. 
  20. ↑ Table D.1 in Brown, Tom B.; Mann, Benjamin; Ryder, Nick; Subbiah, Melanie; Kaplan, Jared; Dhariwal, Prafulla; Neelakantan, Arvind; Shyam, Pranav; Sastry, Girish; Askell, Amanda; Agarwal, Sandhini; Herbert-Voss, Ariel; Krueger, Gretchen; Henighan, Tom; Child, Rewon; Ramesh, Aditya; Ziegler, Daniel M.; Wu, Jeffrey; Winter, Clemens; Hesse, Christopher; Chen, Mark; Sigler, Eric; Litwin, Mateusz; Gray, Scott; Chess, Benjamin; Clark, Jack; Berner, Christopher; McCandlish, Sam; Radford, Alec; Sutskever, Ilya; Amodei, Dario (May 28, 2020). "Language Models are Few-Shot Learners". arXiv:2005.14165v4 [cs.CL].
  21. ↑ "ChatGPT: Optimizing Language Models for Dialogue". 2022-11-30. https://openai.com/blog/chatgpt/. 
  22. ↑ "GPT Neo". March 15, 2023. https://github.com/EleutherAI/gpt-neo. 
  23. ↑ 23.0 23.1 23.2 Gao, Leo; Biderman, Stella; Black, Sid; Golding, Laurence; Hoppe, Travis; Foster, Charles; Phang, Jason; He, Horace; Thite, Anish; Nabeshima, Noa; Presser, Shawn; Leahy, Connor (31 December 2020). "The Pile: An 800GB Dataset of Diverse Text for Language Modeling". arXiv:2101.00027 [cs.CL].
  24. ↑ 24.0 24.1 Iyer, Abhishek (15 May 2021). "GPT-3's free alternative GPT-Neo is something to be excited about". VentureBeat. https://venturebeat.com/ai/gpt-3s-free-alternative-gpt-neo-is-something-to-be-excited-about/. 
  25. ↑ "GPT-J-6B: An Introduction to the Largest Open Source GPT Model | Forefront". https://www.forefront.ai/blog-posts/gpt-j-6b-an-introduction-to-the-largest-open-sourced-gpt-model. 
  26. ↑ 26.0 26.1 26.2 26.3 Dey, Nolan; Gosal, Gurpreet; Zhiming; Chen; Khachane, Hemant; Marshall, William; Pathria, Ribhu; Tom, Marvin; Hestness, Joel (2023-04-01). "Cerebras-GPT: Open Compute-Optimal Language Models Trained on the Cerebras Wafer-Scale Cluster". arXiv:2304.03208 [cs.LG].
  27. ↑ Alvi, Ali; Kharya, Paresh (11 October 2021). "Using DeepSpeed and Megatron to Train Megatron-Turing NLG 530B, the World's Largest and Most Powerful Generative Language Model". https://www.microsoft.com/en-us/research/blog/using-deepspeed-and-megatron-to-train-megatron-turing-nlg-530b-the-worlds-largest-and-most-powerful-generative-language-model/. 
  28. ↑ 28.0 28.1 Smith, Shaden; Patwary, Mostofa; Norick, Brandon; LeGresley, Patrick; Rajbhandari, Samyam; Casper, Jared; Liu, Zhun; Prabhumoye, Shrimai; Zerveas, George; Korthikanti, Vijay; Zhang, Elton; Child, Rewon; Aminabadi, Reza Yazdani; Bernauer, Julie; Song, Xia (2022-02-04). "Using DeepSpeed and Megatron to Train Megatron-Turing NLG 530B, A Large-Scale Generative Language Model". arXiv:2201.11990 [cs.CL].
  29. ↑ 29.0 29.1 Rajbhandari, Samyam; Li, Conglong; Yao, Zhewei; Zhang, Minjia; Aminabadi, Reza Yazdani; Awan, Ammar Ahmad; Rasley, Jeff; He, Yuxiong (2022-07-21), DeepSpeed-MoE: Advancing Mixture-of-Experts Inference and Training to Power Next-Generation AI Scale 
  30. ↑ Wang, Shuohuan; Sun, Yu; Xiang, Yang; Wu, Zhihua; Ding, Siyu; Gong, Weibao; Feng, Shikun; Shang, Junyuan; Zhao, Yanbin; Pang, Chao; Liu, Jiaxiang; Chen, Xuyi; Lu, Yuxiang; Liu, Weixin; Wang, Xi; Bai, Yangfan; Chen, Qiuliang; Zhao, Li; Li, Shiyong; Sun, Peng; Yu, Dianhai; Ma, Yanjun; Tian, Hao; Wu, Hua; Wu, Tian; Zeng, Wei; Li, Ge; Gao, Wen; Wang, Haifeng (December 23, 2021). "ERNIE 3.0 Titan: Exploring Larger-scale Knowledge Enhanced Pre-training for Language Understanding and Generation". arXiv:2112.12731 [cs.CL].
  31. ↑ "Product". https://www.anthropic.com/product. 
  32. ↑ 32.0 32.1 Askell, Amanda; Bai, Yuntao; Chen, Anna; et al. (9 December 2021). "A General Language Assistant as a Laboratory for Alignment". arXiv:2112.00861 [cs.CL].
  33. ↑ Bai, Yuntao; Kadavath, Saurav; Kundu, Sandipan; et al. (15 December 2022). "Constitutional AI: Harmlessness from AI Feedback". arXiv:2212.08073 [cs.CL].
  34. ↑ 34.0 34.1 34.2 Dai, Andrew M; Du, Nan (December 9, 2021). "More Efficient In-Context Learning with GLaM". https://ai.googleblog.com/2021/12/more-efficient-in-context-learning-with.html. 
  35. ↑ "Language modelling at scale: Gopher, ethical considerations, and retrieval". 8 December 2021. https://www.deepmind.com/blog/language-modelling-at-scale-gopher-ethical-considerations-and-retrieval. 
  36. ↑ 36.0 36.1 36.2 Hoffmann, Jordan; Borgeaud, Sebastian; Mensch, Arthur; et al. (29 March 2022). "Training Compute-Optimal Large Language Models". arXiv:2203.15556 [cs.CL].
  37. ↑ 37.0 37.1 37.2 37.3 Table 20 and page 66 of PaLM: Scaling Language Modeling with Pathways
  38. ↑ 38.0 38.1 Cheng, Heng-Tze; Thoppilan, Romal (January 21, 2022). "LaMDA: Towards Safe, Grounded, and High-Quality Dialog Models for Everything". https://ai.googleblog.com/2022/01/lamda-towards-safe-grounded-and-high.html. 
  39. ↑ Thoppilan, Romal; De Freitas, Daniel; Hall, Jamie; Shazeer, Noam; Kulshreshtha, Apoorv; Cheng, Heng-Tze; Jin, Alicia; Bos, Taylor; Baker, Leslie; Du, Yu; Li, YaGuang; Lee, Hongrae; Zheng, Huaixiu Steven; Ghafouri, Amin; Menegali, Marcelo (2022-01-01). "LaMDA: Language Models for Dialog Applications". arXiv:2201.08239 [cs.CL].
  40. ↑ Black, Sidney; Biderman, Stella; Hallahan, Eric (2022-05-01). "GPT-NeoX-20B: An Open-Source Autoregressive Language Model". Proceedings of BigScience Episode #5 – Workshop on Challenges & Perspectives in Creating Large Language Models. Proceedings of BigScience Episode #5 – Workshop on Challenges & Perspectives in Creating Large Language Models. pp. 95–136. https://aclanthology.org/2022.bigscience-1.9/. Retrieved 2022-12-19. 
  41. ↑ 41.0 41.1 41.2 Hoffmann, Jordan; Borgeaud, Sebastian; Mensch, Arthur; Sifre, Laurent (12 April 2022). "An empirical analysis of compute-optimal large language model training". Deepmind Blog. https://www.deepmind.com/blog/an-empirical-analysis-of-compute-optimal-large-language-model-training. 
  42. ↑ Narang, Sharan; Chowdhery, Aakanksha (April 4, 2022). "Pathways Language Model (PaLM): Scaling to 540 Billion Parameters for Breakthrough Performance" (in en). https://ai.googleblog.com/2022/04/pathways-language-model-palm-scaling-to.html. 
  43. ↑ "Democratizing access to large-scale language models with OPT-175B". https://ai.facebook.com/blog/democratizing-access-to-large-scale-language-models-with-opt-175b/. 
  44. ↑ Zhang, Susan; Roller, Stephen; Goyal, Naman; Artetxe, Mikel; Chen, Moya; Chen, Shuohui; Dewan, Christopher; Diab, Mona; Li, Xian; Lin, Xi Victoria; Mihaylov, Todor; Ott, Myle; Shleifer, Sam; Shuster, Kurt; Simig, Daniel; Koura, Punit Singh; Sridhar, Anjali; Wang, Tianlu; Zettlemoyer, Luke (21 June 2022). "OPT: Open Pre-trained Transformer Language Models". arXiv:2205.01068 [cs.CL].
  45. ↑ "metaseq/projects/OPT/chronicles at main · facebookresearch/metaseq" (in en). https://github.com/facebookresearch/metaseq/tree/main/projects/OPT/chronicles. 
  46. ↑ 46.0 46.1 Khrushchev, Mikhail; Vasilev, Ruslan; Petrov, Alexey; Zinov, Nikolay (2022-06-22), YaLM 100B, https://github.com/yandex/YaLM-100B, retrieved 2023-03-18 
  47. ↑ 47.0 47.1 Lewkowycz, Aitor; Andreassen, Anders; Dohan, David; Dyer, Ethan; Michalewski, Henryk; Ramasesh, Vinay; Slone, Ambrose; Anil, Cem; Schlag, Imanol; Gutman-Solo, Theo; Wu, Yuhuai; Neyshabur, Behnam; Gur-Ari, Guy; Misra, Vedant (30 June 2022). "Solving Quantitative Reasoning Problems with Language Models". arXiv:2206.14858 [cs.CL].
  48. ↑ "Minerva: Solving Quantitative Reasoning Problems with Language Models". 30 June 2022. https://ai.googleblog.com/2022/06/minerva-solving-quantitative-reasoning.html. 
  49. ↑ Ananthaswamy, Anil (8 March 2023). "In AI, is bigger always better?". Nature 615 (7951): 202–205. doi:10.1038/d41586-023-00641-w. PMID 36890378. Bibcode: 2023Natur.615..202A. https://www.nature.com/articles/d41586-023-00641-w. Retrieved 9 March 2023. 
  50. ↑ "bigscience/bloom · Hugging Face". https://huggingface.co/bigscience/bloom. 
  51. ↑ Taylor, Ross; Kardas, Marcin; Cucurull, Guillem; Scialom, Thomas; Hartshorn, Anthony; Saravia, Elvis; Poulton, Andrew; Kerkez, Viktor; Stojnic, Robert (16 November 2022). "Galactica: A Large Language Model for Science". arXiv:2211.09085 [cs.CL].
  52. ↑ "20B-parameter Alexa model sets new marks in few-shot learning". 2 August 2022. https://www.amazon.science/blog/20b-parameter-alexa-model-sets-new-marks-in-few-shot-learning. 
  53. ↑ Soltan, Saleh; Ananthakrishnan, Shankar; FitzGerald, Jack; et al. (3 August 2022). "AlexaTM 20B: Few-Shot Learning Using a Large-Scale Multilingual Seq2Seq Model". arXiv:2208.01448 [cs.CL].
  54. ↑ "AlexaTM 20B is now available in Amazon SageMaker JumpStart | AWS Machine Learning Blog". 17 November 2022. https://aws.amazon.com/blogs/machine-learning/alexatm-20b-is-now-available-in-amazon-sagemaker-jumpstart/. 
  55. ↑ 55.0 55.1 55.2 "Introducing LLaMA: A foundational, 65-billion-parameter large language model". Meta AI. 24 February 2023. https://ai.facebook.com/blog/large-language-model-llama-meta-ai/. 
  56. ↑ 56.0 56.1 56.2 "The Falcon has landed in the Hugging Face ecosystem". https://huggingface.co/blog/falcon. 
  57. ↑ "GPT-4 Technical Report". 2023. https://cdn.openai.com/papers/gpt-4.pdf. 
  58. ↑ Schreiner, Maximilian (2023-07-11). "GPT-4 architecture, datasets, costs and more leaked" (in en-US). https://the-decoder.com/gpt-4-architecture-datasets-costs-and-more-leaked/. 
  59. ↑ Dickson, Ben (22 May 2024). "Meta introduces Chameleon, a state-of-the-art multimodal model". VentureBeat. https://venturebeat.com/ai/meta-introduces-chameleon-a-state-of-the-art-multimodal-model/. 
  60. ↑ "chameleon/LICENSE at e3b711ef63b0bb3a129cf0cf0918e36a32f26e2c · facebookresearch/chameleon" (in en). Meta Research. https://github.com/facebookresearch/chameleon/blob/e3b711ef63b0bb3a129cf0cf0918e36a32f26e2c/LICENSE. 
  61. ↑ Dey, Nolan (March 28, 2023). "Cerebras-GPT: A Family of Open, Compute-efficient, Large Language Models". https://www.cerebras.net/blog/cerebras-gpt-a-family-of-open-compute-efficient-large-language-models/. 
  62. ↑ "Abu Dhabi-based TII launches its own version of ChatGPT". https://fastcompanyme.com/news/abu-dhabi-based-tii-launches-its-own-version-of-chatgpt/. 
  63. ↑ Penedo, Guilherme; Malartic, Quentin; Hesslow, Daniel; Cojocaru, Ruxandra; Cappelli, Alessandro; Alobeidli, Hamza; Pannier, Baptiste; Almazrouei, Ebtesam; Launay, Julien (2023-06-01). "The RefinedWeb Dataset for Falcon LLM: Outperforming Curated Corpora with Web Data, and Web Data Only". arXiv:2306.01116 [cs.CL].
  64. ↑ "tiiuae/falcon-40b · Hugging Face". 2023-06-09. https://huggingface.co/tiiuae/falcon-40b. 
  65. ↑ UAE's Falcon 40B, World's Top-Ranked AI Model from Technology Innovation Institute, is Now Royalty-Free , 31 May 2023
  66. ↑ Wu, Shijie; Irsoy, Ozan; Lu, Steven; Dabravolski, Vadim; Dredze, Mark; Gehrmann, Sebastian; Kambadur, Prabhanjan; Rosenberg, David; Mann, Gideon (March 30, 2023). "BloombergGPT: A Large Language Model for Finance". arXiv:2303.17564 [cs.LG].
  67. ↑ Ren, Xiaozhe; Zhou, Pingyi; Meng, Xinfan; Huang, Xinjing; Wang, Yadao; Wang, Weichao; Li, Pengfei; Zhang, Xiaoda; Podolskiy, Alexander; Arshinov, Grigory; Bout, Andrey; Piontkovskaya, Irina; Wei, Jiansheng; Jiang, Xin; Su, Teng; Liu, Qun; Yao, Jun (March 19, 2023). "PanGu-Σ: Towards Trillion Parameter Language Model with Sparse Heterogeneous Computing". arXiv:2303.10845 [cs.CL].
  68. ↑ Köpf, Andreas; Kilcher, Yannic; von Rütte, Dimitri; Anagnostidis, Sotiris; Tam, Zhi-Rui; Stevens, Keith; Barhoum, Abdullah; Duc, Nguyen Minh; Stanley, Oliver; Nagyfi, Richárd; ES, Shahul; Suri, Sameer; Glushkov, David; Dantuluri, Arnav; Maguire, Andrew (2023-04-14). "OpenAssistant Conversations – Democratizing Large Language Model Alignment". arXiv:2304.07327 [cs.CL].
  69. ↑ Wrobel, Sharon. "Tel Aviv startup rolls out new advanced AI language model to rival OpenAI". ISSN 0040-7909. https://www.timesofisrael.com/ai21-labs-rolls-out-new-advanced-ai-language-model-to-rival-openai/. 
  70. ↑ Wiggers, Kyle (2023-04-13). "With Bedrock, Amazon enters the generative AI race". https://techcrunch.com/2023/04/13/with-bedrock-amazon-enters-the-generative-ai-race/. 
  71. ↑ 71.0 71.1 Elias, Jennifer (16 May 2023). "Google's newest A.I. model uses nearly five times more text data for training than its predecessor". CNBC. https://www.cnbc.com/2023/05/16/googles-palm-2-uses-nearly-five-times-more-text-data-than-predecessor.html. 
  72. ↑ "Introducing PaLM 2". May 10, 2023. https://blog.google/technology/ai/google-palm-2-ai-large-language-model/. 
  73. ↑ 73.0 73.1 "Introducing Llama 2: The Next Generation of Our Open Source Large Language Model". 2023. https://ai.meta.com/llama/. 
  74. ↑ "llama/MODEL_CARD.md at main · meta-llama/llama". https://github.com/meta-llama/llama/blob/main/MODEL_CARD.md. 
  75. ↑ "Claude 2". https://www.anthropic.com/index/claude-2. 
  76. ↑ Nirmal, Dinesh (2023-09-07). "Building AI for business: IBM's Granite foundation models" (in en-US). https://www.ibm.com/blog/building-ai-for-business-ibms-granite-foundation-models. 
  77. ↑ "Announcing Mistral 7B". 2023. https://mistral.ai/news/announcing-mistral-7b/. 
  78. ↑ "Introducing Claude 2.1". https://www.anthropic.com/index/claude-2-1. 
  79. ↑ xai-org/grok-1, xai-org, 2024-03-19, https://github.com/xai-org/grok-1, retrieved 2024-03-19 
  80. ↑ "Grok-1 model card". https://x.ai/model-card/. 
  81. ↑ "Gemini – Google DeepMind". https://deepmind.google/technologies/gemini/#capabilities. 
  82. ↑ Franzen, Carl (11 December 2023). "Mistral shocks AI community as latest open source model eclipses GPT-3.5 performance". https://venturebeat.com/ai/mistral-shocks-ai-community-as-latest-open-source-model-eclipses-gpt-3-5-performance/. 
  83. ↑ "Mixtral of experts". 11 December 2023. https://mistral.ai/news/mixtral-of-experts/. 
  84. ↑ AI, Mistral (2024-04-17). "Cheaper, Better, Faster, Stronger". https://mistral.ai/news/mixtral-8x22b/. 
  85. ↑ 85.0 85.1 DeepSeek-AI; Bi, Xiao; Chen, Deli; Chen, Guanting; Chen, Shanhuang; Dai, Damai; Deng, Chengqi; Ding, Honghui et al. (2024-01-05), DeepSeek LLM: Scaling Open-Source Language Models with Longtermism 
  86. ↑ 86.0 86.1 Hughes, Alyssa (12 December 2023). "Phi-2: The surprising power of small language models". https://www.microsoft.com/en-us/research/blog/phi-2-the-surprising-power-of-small-language-models/. 
  87. ↑ "Our next-generation model: Gemini 1.5". 15 February 2024. https://blog.google/technology/ai/google-gemini-next-generation-model-february-2024/#context-window. "This means 1.5 Pro can process vast amounts of information in one go — including 1 hour of video, 11 hours of audio, codebases with over 30,000 lines of code or over 700,000 words. In our research, we’ve also successfully tested up to 10 million tokens." 
  88. ↑ "Gemma". https://ai.google.dev/gemma/terms. 
  89. ↑ "Introducing the next generation of Claude". https://www.anthropic.com/news/claude-3-family. 
  90. ↑ "Databricks Open Model License". 27 March 2024. https://www.databricks.com/legal/open-model-license. 
  91. ↑ "Databricks Open Model Acceptable Use Policy". 27 March 2024. https://www.databricks.com/legal/acceptable-use-policy-open-model. 
  92. ↑ "Fugaku-LLM Terms of Use". 23 April 2024. https://huggingface.co/Fugaku-LLM/Fugaku-LLM-13B/blob/main/LICENSE. 
  93. ↑ "Fugaku-LLM/Fugaku-LLM-13B · Hugging Face". https://huggingface.co/Fugaku-LLM/Fugaku-LLM-13B. 
  94. ↑ "Phi-3". 23 April 2024. https://azure.microsoft.com/en-us/blog/introducing-phi-3-redefining-whats-possible-with-slms. 
  95. ↑ "Phi-3 Model Documentation". https://huggingface.co/docs/transformers/main/en/model_doc/phi3. 
  96. ↑ "Qwen2". https://github.com/QwenLM/Qwen2?spm=a3c0i.28768018.7084722650.1.5cd35c10NEqBXm&file=Qwen1.5. 
  97. ↑ DeepSeek-AI; Liu, Aixin; Feng, Bei; Wang, Bin; Wang, Bingxuan; Liu, Bo; Zhao, Chenggang; Dengr, Chengqi et al. (2024-06-19), DeepSeek-V2: A Strong, Economical, and Efficient Mixture-of-Experts Language Model 
  98. ↑ "NVIDIA Open Models License". 16 June 2025. https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-open-model-license/. 
  99. ↑ "Trustworthy AI". 27 June 2024. https://www.nvidia.com/en-us/agreements/trustworthy-ai/terms/. 
  100. ↑ "nvidia/Nemotron-4-340B-Base · Hugging Face". 2024-06-14. https://huggingface.co/nvidia/Nemotron-4-340B-Base. 
  101. ↑ "Nemotron-4 340B | Research". https://research.nvidia.com/publication/2024-06_nemotron-4-340b. 
  102. ↑ "Introducing Claude 3.5 Sonnet" (in en). https://www.anthropic.com/news/claude-3-5-sonnet. 
  103. ↑ "Introducing computer use, a new Claude 3.5 Sonnet, and Claude 3.5 Haiku" (in en). https://www.anthropic.com/news/3-5-models-and-computer-use. 
  104. ↑ "The Llama 3 Herd of Models" (July 23, 2024) Llama Team, AI @ Meta
  105. ↑ "llama-models/models/llama3_1/MODEL_CARD.md at main · meta-llama/llama-models" (in en). https://github.com/meta-llama/llama-models/blob/main/models/llama3_1/MODEL_CARD.md. 
  106. ↑ "Introducing OpenAI o1". https://openai.com/o1/. 
  107. ↑ 107.0 107.1 "Models Overview". https://docs.mistral.ai/getting-started/models/models_overview/. 
  108. ↑ deepseek-ai/DeepSeek-V3, DeepSeek, 2024-12-26, https://github.com/deepseek-ai/DeepSeek-V3?tab=readme-ov-file, retrieved 2024-12-26 
  109. ↑ Feng, Coco (25 March 2025). "DeepSeek wows coders with more powerful open-source V3 model" (in en). https://www.scmp.com/tech/big-tech/article/3303798/deepseeks-upgraded-foundational-model-excels-coding-and-maths. 
  110. ↑ Amazon Nova Micro, Lite, and Pro - AWS AI Service Cards3, Amazon, 2024-12-27, https://docs.aws.amazon.com/ai/responsible-ai/nova-micro-lite-pro/overview.html, retrieved 2024-12-27 
  111. ↑ deepseek-ai/DeepSeek-R1, DeepSeek, 2025-01-21, https://github.com/deepseek-ai/DeepSeek-R1, retrieved 2025-01-21 
  112. ↑ DeepSeek-AI; Guo, Daya; Yang, Dejian; Zhang, Haowei; Song, Junxiao; Zhang, Ruoyu; Xu, Runxin; Zhu, Qihao et al. (2025-01-22), DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning 
  113. ↑ Qwen; Yang, An; Yang, Baosong; Zhang, Beichen; Hui, Binyuan; Zheng, Bo; Yu, Bowen; Li, Chengyuan et al. (2025-01-03), Qwen2.5 Technical Report 
  114. ↑ 114.0 114.1 MiniMax; Li, Aonian; Gong, Bangwei; Yang, Bo; Shan, Boji; Liu, Chang; Zhu, Cheng; Zhang, Chunhao et al. (2025-01-14), MiniMax-01: Scaling Foundation Models with Lightning Attention 
  115. ↑ MiniMax-AI/MiniMax-01, MiniMax, 2025-01-26, https://github.com/MiniMax-AI/MiniMax-01?tab=readme-ov-file, retrieved 2025-01-26 
  116. ↑ Kavukcuoglu, Koray (5 February 2025). "Gemini 2.0 is now available to everyone". https://blog.google/technology/google-deepmind/gemini-model-updates-february-2025/. 
  117. ↑ "Gemini 2.0: Flash, Flash-Lite and Pro". https://developers.googleblog.com/en/gemini-2-family-expands/. 
  118. ↑ Franzen, Carl (5 February 2025). "Google launches Gemini 2.0 Pro, Flash-Lite and connects reasoning model Flash Thinking to YouTube, Maps and Search". VentureBeat. https://venturebeat.com/ai/google-launches-gemini-2-0-pro-flash-lite-and-connects-reasoning-model-flash-thinking-to-youtube-maps-and-search/. 
  119. ↑ "Claude 3.7 Sonnet and Claude Code" (in en). https://www.anthropic.com/news/claude-3-7-sonnet. 
  120. ↑ "Introducing GPT-4.5". https://openai.com/index/introducing-gpt-4-5/. 
  121. ↑ "Grok 3 Beta — The Age of Reasoning Agents" (in en). https://x.ai/blog/grok-3. 
  122. ↑ Kavukcuoglu, Koray (25 March 2025). "Gemini 2.5: Our most intelligent AI model". https://blog.google/technology/google-deepmind/gemini-model-thinking-updates-march-2025/. 
  123. ↑ "meta-llama/Llama-4-Maverick-17B-128E · Hugging Face". 2025-04-05. https://huggingface.co/meta-llama/Llama-4-Maverick-17B-128E. 
  124. ↑ "The Llama 4 herd: The beginning of a new era of natively multimodal AI innovation" (in en). https://ai.meta.com/blog/llama-4-multimodal-intelligence/. 
  125. ↑ "Introducing OpenAI o3 and o4-mini". https://openai.com/index/introducing-o3-and-o4-mini/. 
  126. ↑ Team, Qwen (2025-04-29). "Qwen3: Think Deeper, Act Faster" (in en). https://qwenlm.github.io/blog/qwen3/. 
  127. ↑ "Introducing Claude 4" (in en). https://www.anthropic.com/news/claude-4. 
  128. ↑ "zai-org/GLM-4.5 · Hugging Face". 2025-08-04. https://huggingface.co/zai-org/GLM-4.5. 
  129. ↑ "GLM-4.5: Reasoning, Coding, and Agentic Abililties" (in en). https://z.ai/blog/glm-4.5. 
  130. ↑ Whitwam, Ryan (5 August 2025). "OpenAI announces two "gpt-oss" open AI models, and you can download them today" (in en). https://arstechnica.com/ai/2025/08/openai-releases-its-first-open-source-models-since-2019/. 
  131. ↑ "Claude Opus 4.1" (in en). https://www.anthropic.com/news/claude-opus-4-1. 
  132. ↑ "Introducing GPT-5". 7 August 2025. https://openai.com/index/introducing-gpt-5/. 
  133. ↑ "OpenAI Platform: GPT-5 Model Documentation". https://platform.openai.com/docs/models/gpt-5. 
  134. ↑ "deepseek-ai/DeepSeek-V3.1 · Hugging Face". 2025-08-21. https://huggingface.co/deepseek-ai/DeepSeek-V3.1. 
  135. ↑ "DeepSeek-V3.1 Release | DeepSeek API Docs" (in en). https://api-docs.deepseek.com/news/news250821. 
  136. ↑ "Introducing Claude Sonnet 4.5" (in en). https://www.anthropic.com/news/claude-sonnet-4-5. 
  137. ↑ "Introducing DeepSeek-V3.2-Exp | DeepSeek API Docs" (in en). https://api-docs.deepseek.com/news/news250929. 
  138. ↑ "deepseek-ai/DeepSeek-V3.2-Exp · Hugging Face". 2025-09-29. https://huggingface.co/deepseek-ai/DeepSeek-V3.2-Exp. 
  139. ↑ "DeepSeek-V3.2-Exp/DeepSeek_V3_2.pdf at main · deepseek-ai/DeepSeek-V3.2-Exp" (in en). https://github.com/deepseek-ai/DeepSeek-V3.2-Exp/blob/main/DeepSeek_V3_2.pdf. 
  140. ↑ "GLM-4.6: Advanced Agentic, Reasoning and Coding Capabilities" (in en). https://z.ai/blog/glm-4.6. 
  141. ↑ "zai-org/GLM-4.6 · Hugging Face". 2025-09-30. https://huggingface.co/zai-org/GLM-4.6. 
  142. ↑ "GLM-4.6". https://modelscope.cn/models/ZhipuAI/GLM-4.6. 



Retrieved from "https://handwiki.org/wiki/index.php?title=Software:List_of_large_language_models&oldid=4139454"