Activefrontierllm Open Source

Qwen3 72B

by Alibaba· Released May 2025· Cutoff April 2025

Qwen3 72B is a large language model from Alibaba's Qwen3 series, featuring a Mixture-of-Experts (MoE) architecture with 72B total parameters and 12B activated parameters. It supports extended context lengths up to 128K tokens and offers strong performance in reasoning, coding, and multilingual tasks.

Official Site API Docs 🤗 Hugging Face

Input cost

Free (open source)

Output cost

Free (open source)

Context window

128K tokens

Max output

8192 tokens

Modalities

text

Parameters

72B (12B activated)

License

Apache-2.0

Capabilities

Function CallingCode GenerationStreamingJSON ModeMultilingual SupportReasoningTool Use

Best For

Complex reasoning, coding, and multilingual tasks requiring a balance of performance and efficiency.

Strengths

Strong reasoning and problem-solving capabilities
Efficient MoE architecture with 12B activated parameters
Supports 128K context window
Multilingual proficiency including English, Chinese, and more

Limitations

Requires significant hardware for full deployment (72B total parameters)
Not multimodal (text-only)
May lag behind larger models in extremely specialized domains

Use Cases

Advanced code generation and debugging

Complex mathematical reasoning

Multilingual translation and content creation

Document analysis with long context

Chatbots and virtual assistants

Data extraction and summarization

Educational tutoring and explanation

Improvements Over Previous Model

Introduced MoE architecture (72B total, 12B activated) vs Qwen2.5 72B dense model
Context window extended from 32K to 128K tokens
Improved reasoning benchmarks (e.g., MATH, GSM8K) by ~5-10%
Better multilingual support with expanded vocabulary
Enhanced tool use and function calling capabilities

Back to all models

Activefrontierllm Open Source

Qwen3 72B

by Alibaba· Released May 2025· Cutoff April 2025

Official Site API Docs 🤗 Hugging Face

Input cost

Free (open source)

Output cost

Free (open source)

Context window

128K tokens

Max output

8192 tokens

Modalities

text

Parameters

72B (12B activated)

License

Apache-2.0

Capabilities

Function CallingCode GenerationStreamingJSON ModeMultilingual SupportReasoningTool Use

Best For

Complex reasoning, coding, and multilingual tasks requiring a balance of performance and efficiency.

Strengths

Strong reasoning and problem-solving capabilities
Efficient MoE architecture with 12B activated parameters
Supports 128K context window
Multilingual proficiency including English, Chinese, and more

Limitations

Requires significant hardware for full deployment (72B total parameters)
Not multimodal (text-only)
May lag behind larger models in extremely specialized domains

Use Cases

Advanced code generation and debugging

Complex mathematical reasoning

Multilingual translation and content creation

Document analysis with long context

Chatbots and virtual assistants

Data extraction and summarization

Educational tutoring and explanation

Improvements Over Previous Model

Introduced MoE architecture (72B total, 12B activated) vs Qwen2.5 72B dense model
Context window extended from 32K to 128K tokens
Improved reasoning benchmarks (e.g., MATH, GSM8K) by ~5-10%
Better multilingual support with expanded vocabulary
Enhanced tool use and function calling capabilities

Back to all models