# Qwen2.5-VL 72B Instruct API prices

Qwen2.5-VL 72B Instruct (qwen) — 131k context, 115k max output.
6 metered per-token offers from 6 provider(s), in USD per 1M tokens.

| Provider | Input $/1M | Output $/1M | Cache read $/1M | Context |
| --- | --- | --- | --- | --- |
| NovitaAI | $0.8 | $0.8 | — | 33k |
| OpenRouter | $0.8 | $1 | $0.4 | 128k |
| OVHcloud AI Endpoints | $1.01 | $1.01 | — | 33k |
| Cortecs | $1.014 | $1.014 | — | 32k |
| Alibaba (China) | $2.294 | $6.881 | — | 131k |
| Alibaba | $2.8 | $8.4 | — | 131k |

## Summary

- Cheapest input: $0.8 per 1M tokens (NovitaAI)
- Cheapest output: $0.8 per 1M tokens (NovitaAI)
- First-party: $8.4 per 1M output tokens (Alibaba)
- Free offers: none
- Subscription plans (not per-token): none
- Released: 2024-09
- Inputs: image, text, video

HTML page: https://llmprice.gitlab.io/models/qwen2-5-vl-72b-instruct/
Full dataset: https://llmprice.gitlab.io/data/catalog.json