Models with recorded evidence.

Start with a research area, drill into a vendor, or page through the full index. Vendor aliases and legacy area IDs are grouped here; original model IDs and model links stay unchanged. Only models with at least one benchmark score appear — a model without a recorded score can’t be ranked.

Vendor:Areas overview SpeakLeash · 263 Alibaba · 104 Google · 102 OpenAI · 86 Meta · 68 Microsoft · 49 Anthropic · 44 DeepSeek · 34 Mistral · 30 mistralai · 19 CYFRAGOVPL · 14 NVIDIA · 14 Zhipu AI · 13 internlm · 10 xAI · 10 ByteDance · 9 Baidu · 8 ibm-granite · 8 PLLuM · 8 allenai · 7 Amazon · 7 MiniMax · 7 Mistral AI · 7 Remek · 7 Shanghai AI Lab · 7 utter-project · 7 CohereForAI · 6 Salesforce · 6 01-ai · 5 Cohere · 5 Moonshot AI · 5 NousResearch · 5 THUML · 5 gguf-iq · 4 IBM · 4 Meituan · 4 openchat · 4 Stanford · 4 THUDM · 4 tiiuae · 4 UC San Diego · 4 VikParuchuri · 4 Allen AI · 3 BAAI · 3 Du et al. · 3 ForgeCode · 3 Fudan University · 3 gguf · 3 gguf11bv30 · 3 gguf7bv30 · 3 IDEA Research · 3 Liao et al. · 3 Moonshot.AI · 3 Nam Tuan Ly / NII · 3 OpenDataLab · 3 OPI-PG · 3 upstage · 3 ViCoS Lab Ljubljana · 3 Xiaomi · 3 Zhao et al. · 3+ 243 smaller vendors (288 models)

§ 01 · Meta models

68 models from Meta · page 1 of 2.

#	Model	Vendor	Parameters	Architecture	Benchmarks	Results
001	meta-llama/Llama-3.3-70B-Instruct	Meta	70.6B	—	4	18
002	meta-llama/Llama-3.2-1B-Instruct	Meta	1.24B	—	3	17
003	meta-llama/Llama-3.2-3B-Instruct	Meta	3.21B	—	3	17
004	meta-llama/Meta-Llama-3.1-70B-Instruct	Meta	70.6B	—	3	17
005	meta-llama/Meta-Llama-3-70B-Instruct	Meta	70.6B	—	3	17
006	meta-llama/Meta-Llama-3-8B-Instruct	Meta	8.03B	—	3	17
007	meta-llama/Llama-4-Scout-17B-16E-Instruct (API)	Meta	109B	—	2	16
008	meta-llama/Meta-Llama-3.1-8B-Instruct	Meta	8.03B	—	2	16
009	Llama 4 Maverick	Meta	400B total / 17B active (128 experts)	Mixture-of-Experts Transformer	9	15
010	Llama 3.1 405B	Meta	—	—	12	13
011	Llama-2-7b-chat-hf	Meta	—	—	1	12
012	Llama-2-7b-hf	Meta	—	—	1	12
013	Meta-Llama-3.1-405B-Instruct-FP8	Meta	—	—	2	12
014	Meta-Llama-3-70B	Meta	—	—	2	12
015	wav2vec 2.0 Large (960h)	Meta	317M	CNN feature encoder + Transformer	9	12
016	Llama-3.2-1B	Meta	—	—	1	11
017	Llama-3.2-3B	Meta	—	—	1	11
018	Llama 3 70B	Meta	—	LLM	11	11
019	Llama-4-Scout-17B-16E	Meta	—	—	1	11
020	Meta-Llama-3.1-70B	Meta	—	—	1	11
021	Meta-Llama-3.1-8B	Meta	—	—	1	11
022	Meta-Llama-3-8B	Meta	—	—	1	11
023	Llama-3.2-1B-Instruct	Meta	—	—	1	9
024	Llama-3.2-3B-Instruct	Meta	—	—	1	9
025	Llama 3 (405B, Instruct)	Meta	—	—	9	9
026	Llama-4-Scout	Meta	109B total / 17B active (16 experts)	Mixture-of-Experts Transformer	3	9
027	Meta-Llama-3.1-405B-Instruct	Meta	—	—	1	9
028	Meta-Llama-3.1-70B-Instruct	Meta	—	—	1	9
029	Meta-Llama-3.1-8B-Instruct	Meta	—	—	1	9
030	DeiT-B	Meta	86M	Vision Transformer	3	8
031	Llama-3.0-70B	Meta	—	—	1	7
032	Llama-3.1-405b	Meta	—	—	1	7
033	Llama-3.1-70B	Meta	—	—	1	7
034	Llama-3.1-8B	Meta	—	—	1	7
035	Llama-3.1-Tulu-3-405B	Meta	—	—	1	7
036	Llama-3.3-70B	Meta	—	—	1	7
037	BART-base (STSM)	Meta	139M	Transformer	1	5
038	meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8 (API)	Meta	402B	—	1	5
039	Llama 3.1 70B	Meta	—	—	4	4
040	Llama-2-70B-chat	Meta	—	Llama 2 70B with RLHF chat fine-tuning	1	3
041	Code Llama 34B	Meta	Unknown	Llama 2 fine-tuned	2	2
042	ConvNeXt V2 Huge	Meta	650M	CNN	2	2
043	DeiT-B Distilled	Meta	86M	Vision Transformer	2	2
044	HuBERT Large (LS-960)	Meta	317M	CNN + Transformer (BERT-style)	1	2
045	Llama 3.2 Vision 90B	Meta	Unknown	Llama 3.1 + cross-attention vision adapter	2	2
046	Muse Spark	Meta	—	—	2	2
047	ViTDet-H (MAE)	Meta	Unknown	Plain ViT-H backbone with simple feature pyramid, Cascade Mask RCNN head	1	2
048	Voicebox	Meta	330M	Flow matching (non-autoregressive)	2	2
049	CodeLlama 70B	Meta	70B	—	1	1
050	convnext_base.fb_in22k_ft_in1k	Meta	—	ConvNeXt-B, IN22K pre-train, IN1K fine-tune	1	1