FlagshipOpen SourceAlibaba

Qwen 3 235B

Qwen 3 235B is Alibaba's flagship open-source language model with 235 billion parameters, featuring a 128K token context window and strong performance in coding and reasoning tasks.

Context 128K
Tier Flagship
Tools Supported
License Open Source
Input from
$0.071 / 1M tokens
across 4 providers

API Pricing

Cheapest on Deep Infra 80% below avg
ProviderInput / 1MOutput / 1MSpeedTTFTUpdated
$0.071$0.10069.9 t/s1.1s4/4/2026
$0.227$0.90669.9 t/s1.1s4/13/2026
$0.455$1.8269.9 t/s1.1s4/14/2026
$0.650$3.0069.9 t/s1.1s4/14/2026

Prices updated daily. Last check: 4/14/2026

Model Details

General

Creator
Alibaba
Family
Qwen
Tier
Flagship
Context Window
128K
Modalities
Text

Capabilities

Tool Calling
Yes
Open Source
Yes
Subtypes
Chat Completion, Code Generation
Aliases
qwen3-235b-a22b, qwen3-235b-a22b-instruct-2507, qwen3-235b-a22b-thinking-2507, qwen3-235b-a22b-2507

Strengths & Limitations

  • Open-source model with full weights available for download and customization
  • Large 235 billion parameter count for complex reasoning and generation tasks
  • 128K token context window supports long-document processing
  • Tool calling functionality with structured output support
  • Strong code generation capabilities across multiple programming languages
  • Benchmark output speed of 65.7 tokens per second for responsive interactions
  • Multilingual support extending beyond English for global applications
  • Significant computational requirements due to 235B parameter size
  • Text-only modality without image or audio input support
  • Time to first token of 1,037ms slower than smaller models
  • Smaller context window compared to some competing flagship models
  • Inference costs scale with model size for cloud deployment

Key Features

235 billion parameter transformer architecture
128,000 token context window
Tool calling with structured JSON output
Code generation and debugging capabilities
Chat completion with multi-turn conversation support
Open-source licensing with downloadable model weights
Multilingual text processing and generation
Streaming response generation

About Qwen 3 235B

Qwen 3 235B is Alibaba's flagship language model in the Qwen family, representing their largest open-source offering with 235 billion parameters. As the top-tier model in the Qwen 3 series, it targets complex reasoning, coding, and chat completion tasks where maximum capability is required over efficiency considerations. The model operates with a 128,000 token context window and focuses on text-based tasks including chat completion and code generation. It includes tool calling capabilities and demonstrates benchmark performance of 65.7 output tokens per second with a time to first token of 1,037 milliseconds. The model builds on Alibaba's continued development of the Qwen architecture with improvements in reasoning and multilingual capabilities. Qwen 3 235B serves applications requiring sophisticated language understanding and generation, particularly in enterprise environments where open-source licensing provides deployment flexibility. Its substantial parameter count positions it among the larger open-source models available, though users must balance its capabilities against the computational resources required for inference.

Common Use Cases

Qwen 3 235B is designed for demanding applications that require maximum language model capability, including complex reasoning tasks, advanced code generation and debugging, long-document analysis within its 128K context window, and sophisticated chatbot implementations. Its open-source nature makes it particularly suitable for organizations requiring on-premises deployment, model customization, or fine-tuning for specialized domains. The model's substantial size makes it appropriate for use cases where response quality is prioritized over inference speed, such as research applications, content creation workflows, and enterprise assistant implementations that benefit from its tool calling capabilities.

Frequently Asked Questions

How much does Qwen 3 235B cost per million tokens?

Qwen 3 235B pricing varies by provider and may include both hosted API access and self-hosting options since it's open-source. Check the pricing table above for current rates across all providers offering this model.

What is Qwen 3 235B best used for?

Qwen 3 235B excels at complex reasoning tasks, advanced code generation and debugging, long-document processing up to 128K tokens, and applications requiring tool calling capabilities. Its 235B parameter size makes it suitable for demanding use cases where response quality is more important than speed.

Can I run Qwen 3 235B on my own infrastructure?

Yes, Qwen 3 235B is open-source with downloadable model weights, allowing self-hosting and customization. However, the 235 billion parameter size requires significant computational resources including multiple high-end GPUs and substantial memory for efficient inference.