Together AI GPU pricing

Together AI lists 4 GPU models across 8 configurations, starting at $1.99 per GPU-hour for the H100 SXM. It is the cheapest dedicated provider we track for 2 of its 8 GPU-and-pricing-model combinations: B200 (on-demand), B300 (on-demand).

Every comparison below is like-for-like: same GPU, same pricing model, and same capacity tier. Together AI is dedicated capacity, so it is measured against the other 35 dedicated providers on the table — not against peer marketplace listings, which are usually cheaper and are not equivalent.

Together AI prices vs. the cheapest dedicated provider

GPUVRAMPricingTogether AI $/GPU-hr Best dedicatedDifferenceCheapest at
H100 SXM 80 GB on-demand $1.99 $1.79 +11% Cudo Compute rent
H200 141 GB on-demand $2.99 $2.91 +3% Seeweb rent
H100 SXM 80 GB reserved $3.99 $2.28 +75% LeaderGPU rent
B200 192 GB on-demand $4.09 $4.09 cheapest — rent
B300 288 GB on-demand $4.99 $4.99 cheapest — rent
H200 141 GB reserved $5.99 $2.99 +100% fal rent
B200 192 GB reserved $8.19 $5.49 +49% fal rent
B300 288 GB reserved $9.99 $5.99 +67% fal rent

Questions

How much does a GPU cost on Together AI?
Together AI starts at $1.99 per GPU-hour for the H100 SXM. It lists 4 GPU models in total: B200, B300, H100 SXM, H200.

Is Together AI cheaper than other GPU clouds?
For 2 of the 8 GPU and pricing-model combinations it offers, Together AI is the cheapest dedicated provider on this table — including B200, B300. For the rest, the table above names the provider that beats it and by how much. Comparisons are like-for-like on GPU, pricing model and capacity tier.

Which GPUs does Together AI offer?
4 models: B200, B300, H100 SXM, H200. Availability changes; the "in stock" flags on each GPU page reflect the last scrape.

Compare with other providers

Browse by GPU

All providers · All GPUs · Live sortable table