The Orchestrator › Models › Qwen
Qwen: Qwen3 VL 8B Instruct
Qwen3-VL-8B-Instruct is a multimodal vision-language model from the Qwen3-VL series, built for high-fidelity understanding and reasoning across text, images, and video. It features improved multimodal fusion with Interleaved-MRoPE for long-horizon...
Specifications
| Provider | Qwen |
| Model ID | qwen/qwen3-vl-8b-instruct |
| Context window | 262K tokens (262,144) |
| Modalities | text + image → text |
| Tokenizer | Qwen3 |
| Router sources | OpenRouter, TrustedRouter |
| Released | 2025-10-14 |
| Tracked since | 2026-11-14 |
| Popularity | Not in the ranked set |
| Reasoning | — |
Pricing (per million tokens)
| Variant | Source | API ID | Input | Output | Context |
|---|---|---|---|---|---|
| standard | OpenRouter | qwen/qwen3-vl-8b-instruct | $0.25 | $0.75 | 262K |
| standard | TrustedRouter | qwen/qwen3-vl-8b-instruct | $0.192 | $0.791 | 262K |
Recorded price history
| Date | Input / 1M | Output / 1M | Context |
|---|---|---|---|
| 2026-11-14 | $0.192 | $0.791 | 262K |
| 2026-11-14 | $0.25 | $0.75 | 262K |
Capabilities
Vision (image input)Agentic tool useStructured outputs256K+ context windowNo content storage (attested)Zero data retention
Trust flags come from TrustedRouter's per-provider attestations.
Supported parameters
frequency_penalty, logit_bias, logprobs, max_tokens, presence_penalty, repetition_penalty, response_format, seed, stop, structured_outputs, temperature, tool_choice, tools, top_k, top_logprobs, top_p