Qwen2.5-VL 72B Instruct vs minicpm-v-4.5: API price and specs

Qwen2.5-VL 72B Instruct is 23% cheaper than minicpm-v-4.5 on input price.

Qwen2.5-VL 72B Instruct is 55% cheaper than minicpm-v-4.5 on output price.

Qwen2.5-VL 72B Instruct is 31% cheaper than minicpm-v-4.5 on a 1M-in / 200k-out workload.

Qwen2.5-VL 72B Instruct has a 300% larger context than minicpm-v-4.5.

Field	Qwen2.5-VL 72B Instruct	minicpm-v-4.5
Lowest input, per 1M	$0.50	$0.65 +30.6%
Median input, per 1M	$1.01 +53.5%	$0.66
Lowest output, per 1M	$0.50	$1.10 +120.2%
Median output, per 1M	$1.01	$1.11 +9.9%
Lowest unit price	—	—
Cost of 1M in + 200k out	$0.60	$0.87 +45.5%
Context	128K	32K +75%
Max output	115K	32K +72.2%
Providers	10	3 +70%
Regions	EU	GLOBAL
Accepts	image, image/bmp, image/heic, image/jpeg, image/png, image/tiff, image/webp, text, video, video/mp4, video/quicktime, video/x-flv, video/x-matroska, video/x-ms-wmv, video/x-msvideo	image, text
Returns	text	text
Type	Text models	Text models
Free option	no	no
Cheapest provider	aihubmix	cortecs

Your own workload: how many tokens the model gets (input) and gives back (output) - for example, in a month. The "Cost" row of the table above is recalculated for these numbers, for each model.

Computed from current offers each time the page opens; the best value of a row is bold, the percent is how far behind the best.

Providers selling both

cortecs, edenai, trustedrouter

Frequently asked questions

Which is cheaper, Qwen2.5-VL 72B Instruct or minicpm-v-4.5?

Qwen2.5-VL 72B Instruct is 23% cheaper than minicpm-v-4.5 on input price. Qwen2.5-VL 72B Instruct is 55% cheaper than minicpm-v-4.5 on output price. Qwen2.5-VL 72B Instruct is 31% cheaper than minicpm-v-4.5 on a 1M-in / 200k-out workload.

Which has the larger context, Qwen2.5-VL 72B Instruct or minicpm-v-4.5?

Qwen2.5-VL 72B Instruct: 128K tokens; minicpm-v-4.5: 32K tokens.

Where can I use both Qwen2.5-VL 72B Instruct and minicpm-v-4.5?

cortecs, edenai, trustedrouter