Major LLMs + Data Availability

	A	B	C	D	E	F	G	H	I	J	K	L	M	N
1	Creator	Model	Size	Playground	OSS?	Inference API	Pretraining Corpus	Corpus Public	Finetuning?	Finetuning Corpora?	Fully Documented?	Notes	Inference Link	Training Link
2	OpenAI	davinci	175B?	https://beta.openai.com/playground	No	API	CC + Webtext + Book Corpora + English Wikipedia	Components Yes	None	Proprietary	Partially through blogs/paper		https://openai.com/api/
3		text-davinci-001	175B?	https://beta.openai.com/playground	No	API	CC + Webtext + Book Corpora + English Wikipedia	Components Yes	Instruction tuning	Proprietary	Partially through blogs/paper		https://openai.com/api/
4		code-davinci-002	175B?	https://beta.openai.com/playground	No	API	CC + Webtext + Book Corpora + English Wikipedia	Components Yes	Code tuning	Proprietary	Partially through blogs/paper		https://openai.com/api/
5		text-davinci-002	175B?	https://beta.openai.com/playground	No	API	CC + Webtext + Book Corpora + English Wikipedia	Components Yes	Code + instruction tuning	Proprietary	Partially through blogs/paper		https://openai.com/api/
6		text-davinci-003	175B?	https://beta.openai.com/playground	No	API	CC + Webtext + Book Corpora + English Wikipedia	Components Yes	Code + instructions + RLHF	Proprietary	Partially through blogs/paper	Closest to ChatGPT	https://openai.com/api/
7	Google	LaMDA	136B	https://blog.google/technology/ai/join-us-in-the-ai-test-kitchen/	No	No	Proprietary	No	Dialog tuning	Proprietary	Partially through blogs/paper
8		Flan-LaMDA	136B	No	No	No	Proprietary	No	Instruction tuning	https://github.com/google-research/FLAN/tree/main/flan/	Partially through blogs/paper
9		T5	11B	No	Yes	Sort of (very slow/unreliable through HuggingFace Inference API)	CC	Yes	None	None	Yes, paper
10		UL2	20B	No	Yes	HF (?)	CC	Yes	None	None	Yes, paper
11		Flan-T5	11B	https://huggingface.co/spaces/osanseviero/i-like-flan	Yes	Sort of (very slow/unreliable through HuggingFace Inference API)	CC	Yes	Instruction tuning	Public but not aggregated	Yes, paper
12		PaLM	540B	No	No	No	Proprietary	No	None	None	Partially through blogs/paper
13		U-PaLM	540B	No	No	No	Proprietary	No	None	None	Partially through blogs/paper
14		Flan-PaLM	540B	No	No	No	Proprietary	No	Instruction tuning	Public but not aggregated	Partially through blogs/paper	Current NLP SOTA
15		Flan-U-PaLM	540B	No	No	No	Proprietary	No	Instruction tuning	Public but not aggregated	Partially through blogs/paper	Current NLP SOTA
16	BigScience	BLOOM	175B	https://huggingface.co/bigscience/bloom	Yes	Sort of (very slow/unreliable through HuggingFace Inference API)	ROOTS	Yes	None	None	Yes, paper		https://huggingface.co/inference-api
17		BLOOMZ	175B	https://huggingface.co/bigscience/bloomz	Yes	Sort of (very slow/unreliable through HuggingFace Inference API)	ROOTS	Yes	Instruction tuning	Public collection	Yes, paper	Multilingual	https://huggingface.co/inference-api
18		mT0	11B	https://huggingface.co/bigscience/mt0-xxl	Yes	Sort of (very slow/unreliable through HuggingFace Inference API)	ROOTS	Yes	Instruction tuning	Public collection	Yes, paper	Multilingual, better than BLOOMZ	https://huggingface.co/inference-api
19	Meta	OPT	175B	https://opt.alpa.ai/	Yes	No	The Pile + Pushift Reddit + RoBERTA (BookCorpus + English Wiki)		None	None	Yes, paper		https://github.com/facebookresearch/metaseq/tree/main/projects/OPT
20		OPT-IML	175B	No	Yes	No	The Pile + Pushift Reddit + RoBERTA (BookCorpus + English Wiki)		Instruction tuning	Public but not aggregated	Yes, paper
21		Galactica									Yes, paper
22	Eleuther	GPT-J	6B	https://goose.ai/playground	Yes	Yes	The Pile	Yes	No?	None	Yes, git repo		https://goose.ai/docs/api	https://github.com/EleutherAI/pythia
23		GPT-NeoX	20B	https://goose.ai/playground	Yes	Yes	The Pile	Yes	No?	None	Yes, git repo		https://goose.ai/docs/api	https://github.com/EleutherAI/pythia
24	Cohere	xlarge v20220609	52B	https://os.cohere.ai/playground/xlarge/generate		API	??	??	??	??	No		https://docs.cohere.ai/reference/generate
25	AnthropicAI	Claude
26	Tsinghua University	GLM-130B	130B	No	Yes	No?	??	??	Instruction tuning	Public: Promptsource + DeepStruct	Very well documented in paper and github repo	Better than davinci-001, worse than Flan-T5	https://github.com/THUDM/GLM-130B	https://github.com/THUDM/GLM-130B
27	LLAMA
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100