Catalog

Models

54 variants

All model variants

Variant	Family	Provider	Params (B)	Active (B)	Type	Modalities
DeepSeek-R1	DeepSeek R1	DeepSeek	—	—	reasoning	text
DeepSeek-R1-Distill-Qwen-32B	DeepSeek R1	DeepSeek	32	—	distill	text
DeepSeek-V3-0324	DeepSeek V3	DeepSeek	671	37	moe	text
DeepSeek-V3.2	DeepSeek V3	DeepSeek	—	—	moe	text
Devstral-2-123B-Instruct-2512	Devstral	Mistral AI	123	—	coding	text
Devstral-Small-2507	Devstral	Mistral AI	—	—	coding	text
Kimi-K2-Instruct	Kimi K2	Moonshot AI	1000	32	instruct	text
Kimi-K2-Thinking	Kimi K2	Moonshot AI	1000	32	reasoning	text
Kimi-K2.5	Kimi K2.5	Moonshot AI	—	—	multimodal	text+vision
Kimi-VL-A3B-Instruct	Kimi VL	Moonshot AI	—	—	vision_language	text+vision
Llama-4-Maverick-17B-128E-Instruct	Llama 4	Meta	402	17	instruct	text+vision
Llama-4-Scout-17B-16E-Instruct	Llama 4	Meta	109	17	instruct	text+vision
Magistral-Small-2506	Magistral	Mistral AI	24	—	reasoning	text
Magistral-Small-2509	Magistral	Mistral AI	24	—	reasoning	text
MiniMax-M2	MiniMax M2	MiniMax	230	10	moe	text
MiniMax-M2.5	MiniMax M2	MiniMax	230	10	moe	text
Ministral-3-8B-Base-2512	Ministral 3	Mistral AI	8	—	base	text+vision
Ministral-3-8B-Instruct-2512	Ministral 3	Mistral AI	8	—	instruct	text+vision
Ministral-3-8B-Reasoning-2512	Ministral 3	Mistral AI	8	—	reasoning	text+vision
Mistral-Small-3.1-24B-Instruct-2503	Mistral Small	Mistral AI	24	—	instruct	text+vision
NVIDIA-Nemotron-3-Nano-30B-A3B-NVFP4	Nemotron 3	NVIDIA	30	3	moe	text
NVIDIA-Nemotron-3-Super-120B-A12B-BF16	Nemotron 3	NVIDIA	124	12	moe	text
NVIDIA-Nemotron-3-Super-120B-A12B-Base-BF16	Nemotron 3	NVIDIA	124	12	base	text
OLMo-2-0325-32B	OLMo 2	AllenAI	32	—	base	text
OLMo-2-0325-32B-Instruct	OLMo 2	AllenAI	32	—	instruct	text
OLMo-2-0425-1B-SFT	OLMo 2	AllenAI	1	—	instruct	text
OLMo-2-1124-13B-Instruct	OLMo 2	AllenAI	13	—	instruct	text
OLMo-2-1124-7B-Instruct	OLMo 2	AllenAI	7	—	instruct	text
Phi-4-reasoning	Phi-4	Microsoft	14	—	reasoning	text+vision+audio
Qwen3-14B	Qwen3	Qwen	14	—	instruct	text+vision
Qwen3-235B-A22B	Qwen3	Qwen	235	22	moe	text
Qwen3-32B	Qwen3	Qwen	32	—	instruct	text+vision
Qwen3-8B	Qwen3	Qwen	8	—	instruct	text+vision
Qwen3-VL-32B-Instruct	Qwen3-VL	Qwen	32	—	vision_language	text+vision
Qwen3-VL-8B-Instruct	Qwen3-VL	Qwen	8	—	vision_language	text+vision
Qwen3.5-122B-A10B	Qwen3.5	Qwen	125	10	moe	text
Qwen3.5-27B	Qwen3.5	Qwen	27	—	instruct	text
Qwen3.5-397B-A17B	Qwen3.5	Qwen	403	17	moe	text
gemma-3-12b-it	Gemma 3	Google	12	—	instruct	text+vision
gemma-3-27b-it	Gemma 3	Google	27	—	instruct	text+vision
gemma-3-4b-it	Gemma 3	Google	4	—	instruct	text+vision
gemma-3n-E2B-it	Gemma 3n	Google	—	—	instruct	text+vision
gemma-3n-E4B-it	Gemma 3n	Google	—	—	instruct	text+vision
gpt-oss-120b	GPT-OSS	OpenAI	117	5.1	instruct	text
gpt-oss-20b	GPT-OSS	OpenAI	21	3.6	instruct	text
granite-3.3-2b-base	Granite 3.3	IBM	2	—	base	text
granite-3.3-2b-instruct	Granite 3.3	IBM	2	—	instruct	text
granite-3.3-8b-base	Granite 3.3	IBM	8	—	base	text
granite-3.3-8b-instruct	Granite 3.3	IBM	8	—	instruct	text
granite-vision-3.3-2b	Granite 3.3	IBM	2	—	vision_language	text+vision
phi-4	Phi-4	Microsoft	14	—	base	text
phi-4-mini-instruct	Phi-4	Microsoft	—	—	instruct	text
phi-4-multimodal-instruct	Phi-4	Microsoft	—	—	multimodal	text+vision+audio
phi-4-reasoning-vision-15B	Phi-4	Microsoft	15	—	reasoning	text+vision