A
Qwen3 VL 32B InstructOPEN
Qwen3-VL-32B-Instruct is a large-scale multimodal vision-language model designed for high-precision understanding and reasoning across text, images, and video.
Prices update automatically — checked daily against provider list prices.
- $0.18 per 1M tokens, assuming 3 input for 1 output the convention used by Artificial Analysis
- cheaper than 62% of the 203 open-weight models tracked here
- 9 of 53 Alibaba models tracked here, from cheapest
Benchmarks & performance
Independent benchmark scores for Qwen3 VL 32B Instruct, measured by Artificial Analysis. Higher is better.
For the same budget, Qwen3 VL 32B Instruct delivers 4.2 times more capability. Open weights also mean you can self-host it and pay nothing per token.
Models at this level cost $0.4 per 1M tokens (median of 60) — this one costs $0.18.
Cheaper and better on this index: Claude Haiku 5.5 (batch) · GLM 5.3 Flash (batch) · DeepSeek V4.1 Flash (batch) and 28 more
About this model
Qwen3 VL 32B Instruct is an open-weight AI model by Alibaba. You can download and self-host it for free; the prices below are hosted-API list prices, tracked daily, for when you prefer convenience over self-hosting.
Frequently asked questions
What is Qwen3 VL 32B Instruct?
Qwen3 VL 32B Instruct is an AI language model from Alibaba. It is open-weight: you can download it and run it on your own hardware, for free. It scores 8.4 on the Artificial Analysis intelligence index.
Is Qwen3 VL 32B Instruct free?
The weights are free and open — you can self-host Qwen3 VL 32B Instruct and pay nothing per token. If you prefer a hosted API, list prices are $0.1 per million input tokens and $0.42 per million output tokens.
What is Qwen3 VL 32B Instruct good at?
Independent benchmarks from Artificial Analysis give it GPQA 67.1%, MMLU-Pro 79.1%, Humanity's Last Exam 6.8%, LiveCodeBench 51.4%, AIME 2025 68.3%, IFBench 39.2%, τ²-Bench 29.2%, Terminal-Bench Hard 8.3%. It is particularly used for mathematical reasoning.
Can I self-host Qwen3 VL 32B Instruct?
Yes. Qwen3 VL 32B Instruct has open weights, so you can download it and run it on your own GPU or server with tools like Ollama, vLLM or llama.cpp — with no per-token cost.