Lab

Meta

Lab summary is pending review.

Llama-3.1-70B-Instruct

unknown

Open Llama instruction model for multilingual chat, reasoning, and coding

Context
128,000
Output
4,096
Offers
0
View model

Llama-3.1-8B-Instruct

unknown

Compact open Llama model for lightweight chat, drafting, and self-hosting

Context
128,000
Output
4,096
Offers
0
View model

Llama-3.2-11B-Vision-Instruct

unknown

Open multimodal Llama model for image understanding, captioning, and visual QA

Context
128,000
Output
4,096
Offers
0
View model

Llama-3.2-1B

unknown

Compact open Llama base model for lightweight and on-device use

Context
131,072
Output
8,192
Offers
0
View model

Llama 3.2 1B Instruct

unknown

Compact instruction-tuned Llama model for multilingual dialogue and on-device assistants

Context
131,072
Output
32,000
Offers
0
View model

Llama-3.2-3B

unknown

Small open Llama base model for lightweight text generation and self-hosting

Context
131,072
Output
8,192
Offers
0
View model

Llama 3.2 3B Instruct

unknown

Small instruction-tuned Llama model for multilingual dialogue and agentic retrieval

Context
131,072
Output
131,072
Offers
0
View model

Llama-3.2-90B-Vision-Instruct

unknown

Large open multimodal Llama model for image reasoning, document QA, and captioning

Context
128,000
Output
4,096
Offers
0
View model

Llama-3.3-70B-Instruct

unknown

Popular open Llama workhorse for multilingual chat, coding, and self-hosting

Context
128,000
Output
4,096
Offers
22
View model

Llama 4 Maverick 17B Instruct

unknown

Open multimodal Llama for strong reasoning with efficient everyday serving

Context
1,000,000
Output
16,384
Offers
1
View model

Llama 4 Scout 17B Instruct

unknown

Open Llama with long-context vision for efficient multimodal agents

Context
10,000,000
Output
16,384
Offers
1
View model

Llama-Guard-3-8B

unknown

Llama 3.1-based safety classifier for moderating prompts and model responses

Context
128,000
Output
4,096
Offers
0
View model

Llama Guard 4 12B

unknown

Multimodal Llama 4-based safety classifier for moderating prompts and responses

Context
163,840
Output
163,840
Offers
0
View model

Llama Prompt Guard 2 22M

unknown

22M-parameter (DeBERTa-xsmall) English classifier that detects prompt injection and jailbreak attempts

Context
512
Output
512
Offers
0
View model

Llama Prompt Guard 2 86M

unknown

86M-parameter (mDeBERTa-base) multilingual classifier that detects prompt injection and jailbreak attempts

Context
512
Output
512
Offers
0
View model

Muse Ember

unknown

Muse Ember is Meta's smallest and most efficient reasoning model, built for the work agents repeat all day: tool calls, coding subtasks, routing, extraction, and long-document parsing. It pairs with Muse Wonder or Muse Spark, which plan while Muse Ember handles the volume.

Context
1,048,576
Output
524,288
Offers
0
View model

Muse Glimmer 30B

unknown

Muse Glimmer is a 30-billion-parameter open-weight multimodal model from Meta Superintelligence Labs, distilled from Muse Spark for always-on local agents, tool use, coding, and image understanding.

Context
131,072
Output
131,072
Offers
0
View model

Muse Spark 1.1

unknown

Muse Spark is a natively multimodal reasoning model with support for tool-use, visual chain of thought, and multi-agent orchestration.

Context
1,048,576
Output
131,072
Offers
6
View model

Muse Spark 1.2

unknown

Muse Spark 1.2 is a coding-focused update to Muse Spark 1.1 with improvements in code generation, complex debugging, codebase understanding, and end-to-end developer workflows.

Context
1,048,576
Output
131,072
Offers
0
View model

Muse Spark 1.3

unknown

Muse Spark 1.3 is a multimodal reasoning model from Meta for long-running agentic, multi-agent, and coding workflows. It improves long-horizon agent collaboration, instruction following, and coding efficiency relative to Muse Spark 1.2.

Context
1,048,576
Output
131,072
Offers
0
View model

Muse Spark Open

unknown

Muse Spark Open is an 816B total / 42B active open-weight multimodal reasoning model from Meta for coding and long-horizon agentic work, with 1M-token context.

Context
1,048,576
Output
131,072
Offers
0
View model

Muse Wonder

unknown

Muse Wonder is a frontier multimodal reasoning model from Meta for personal and professional agents, long-horizon coding, and computer use. It combines Active Memory, which carries relevant context across conversations, with Active Focus, which keeps it working through blockers toward a goal.

Context
1,048,576
Output
131,072
Offers
0
View model

Muse Wonder Full

unknown

Muse Wonder Full is the most capable model in Meta's Muse Wonder series, built for long-running professional agents and ultra-long-horizon software engineering. It shares Muse Wonder's Active Memory and Active Focus capabilities and generally scores higher on Meta's agentic and coding evaluations.

Context
1,048,576
Output
Unknown
Offers
0
View model