GPT OSS 120B vs GPT OSS 20B: API price and specs

GPT OSS 20B is 67% cheaper than GPT OSS 120B on input price.

GPT OSS 20B is 59% cheaper than GPT OSS 120B on output price.

GPT OSS 20B is 64% cheaper than GPT OSS 120B on a 1M-in / 200k-out workload.

Field	GPT OSS 120B	GPT OSS 20B
Lowest input, per 1M	$0.027 +200%	$0.009
Median input, per 1M	$0.15 +200%	$0.05
Lowest output, per 1M	$0.098 +145%	$0.04
Median output, per 1M	$0.60 +200%	$0.20
Lowest unit price	—	—
Cost of 1M in + 200k out	$0.0466 +174.1%	$0.017
Context	131K	131K
Max output	131K	131K
Providers	43	28 +34.9%
Regions	EU, GLOBAL, US	EU, GLOBAL, US
Accepts	file, text	file, image, text
Returns	text	text
Type	Text models	Text models
Free option	yes	yes
Cheapest provider	kilo	edenai

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

aihubmix, aiml, cloudflare-workers-ai, cometapi, cortecs, deepinfra, digitalocean, edenai, fastrouter, google-vertex, helicone, huggingface, impossibl, io-net, jiekou, kenari, kilo, ollama, ollama-cloud, openrouter, ovhcloud, poe, regolo-ai, requesty, siliconflow, trustedrouter, vercel

Frequently asked questions

Which is cheaper, GPT OSS 120B or GPT OSS 20B?

GPT OSS 20B is 67% cheaper than GPT OSS 120B on input price. GPT OSS 20B is 59% cheaper than GPT OSS 120B on output price. GPT OSS 20B is 64% cheaper than GPT OSS 120B on a 1M-in / 200k-out workload.

Which has the larger context, GPT OSS 120B or GPT OSS 20B?

GPT OSS 120B: 131K tokens; GPT OSS 20B: 131K tokens.

Where can I use both GPT OSS 120B and GPT OSS 20B?