# Qwen3-VL 4B Instruct Qwen3-VL 4B Instruct by Alibaba Cloud: no offer with a token price is listed at present. Profile and benchmarks are on this page. - [Compare offers](https://models.searchit-agent.ai/ui-fat/modelle_vision?q=qwen3-vl-4b-instruct) - [View price history](https://models.searchit-agent.ai/ui-fat/modelle_vision?q=qwen3-vl-4b-instruct) ## Profile | Profile | | |---|---| | Developer | Alibaba Cloud | | Type | Image models | | Origin | CN | | Input types | text+bild+video | | Context (tokens) | 262,144 | | Max. output (tokens) | – | | Reasoning | ohne | | Parameters (bn) | 4.44 | | Open weights | ✓ yes | | Licence | Apache-2.0 | | VRAM at 4-bit (GB) | 7 (RTX PRO 4000) | | Released | – | [Open in the app](https://models.searchit-agent.ai/ui-fat/modelle_vision?q=qwen3-vl-4b-instruct) ## Offers with processing in the EU (0) No offers in this group. [Open in the app](https://models.searchit-agent.ai/ui-fat/modelle_vision?q=qwen3-vl-4b-instruct) ## Offers outside the EU (0) No offers in this group. [Open in the app](https://models.searchit-agent.ai/ui-fat/modelle_vision?q=qwen3-vl-4b-instruct) ## AI Act Whether the developer has signed the Code of Practice for general-purpose AI models: unknown. What this means is explained under “EU and AI Act”. [EU processing and AI Act](https://models.searchit-agent.ai/en/knowledge/eu-and-ai-act/) ## Benchmarks | Benchmarks | | |---|---:| | GPQA Diamond (Artificial Analysis) | 37.1 | | Humanity's Last Exam (Artificial Analysis) | 3.6 | | HallusionBench (llm-stats-Abzug) | 0.58 | | OCRBench v2, englische Spur (llm-stats-Abzug) | 0.64 | _[Report an error on this page](https://models.searchit-agent.ai/en/report/report-error/?seite=https%3A%2F%2Fmodels.searchit-agent.ai%2Fen%2Fmodels%2Fqwen3-vl-4b-instruct%2F)_ --- Prices as of 14 Oct 2026. Providers' list prices, without guarantee.