Qwen2-VL-7B
Vision
Smaller Qwen2-VL, single-GPU vision model. Cheap self-hosted alternative to GPT-4o-mini for vision tasks.
Qwen2-VL-7B is a multimodal AI model from Qwen. It costs $0.300 per million input tokens and $0.300 per million output tokens (blended $0.300/M), with a 32,000-token context window.
INPUT
$0.300/M
per million input tokens
OUTPUT
$0.300/M
per million output tokens
CONTEXT
32,000
tokens
What it is good at
- Single-GPU vision
- Open weights
- Document understanding
Typical use cases
- On-prem vision QA
- Cheap multimodal RAG
Benchmarks
vs. best public score
Hand-curated from each provider's published reports and public leaderboards. Methodology varies across sources, treat as directional rather than authoritative.
More from Qwen
See all 22 →Qwen3.7-Max
Flagship · 1,000,000 ctx
in $2.500/Mout $7.500/M
Qwen3.7-Plus
Balanced · 1,000,000 ctx
in $0.400/Mout $1.600/M
Qwen2.5-Max
Flagship · 32,000 ctx
in $1.600/Mout $6.400/M
Qwen2.5-72B-Instruct
Large · 128,000 ctx
in $1.200/Mout $1.200/M
Qwen2.5-32B-Instruct
Balanced · 128,000 ctx
in $0.700/Mout $0.700/M
Qwen2.5-14B-Instruct
Small · 128,000 ctx
in $0.350/Mout $0.350/M
Frequently asked questions
How much does Qwen2-VL-7B cost?
Qwen2-VL-7B costs $0.300 per million input tokens and $0.300 per million output tokens, for a blended reference rate of $0.300 per million tokens.
What is Qwen2-VL-7B's context window?
Qwen2-VL-7B supports up to 32,000 tokens of context in a single request.
What is Qwen2-VL-7B best for?
Qwen2-VL-7B is well suited to Single-GPU vision, Open weights and Document understanding.
Who makes Qwen2-VL-7B?
Qwen2-VL-7B is developed and served by Qwen. It was released in Aug 2024.