GPU cloud · 4 GPU models tracked
Together AI
$3.19–$9.99per GPU-hour, the cheapest row to the dearest
GPU cluster rates per GPU-hour: on-demand, and a reserved rate whose term the page does not state. Per-token inference pricing is a different product and not listed.
As of , Together AI lists 4 GPU models we track, from $3.19 per GPU-hour (NVIDIA H100 SXM, reserved, from) to $9.99 (NVIDIA B300, on-demand).
| Website | www.together.ai/ |
|---|---|
| Pricing page | www.together.ai/gpu-clusters |
| Regions and tiers | No region or tier named in the rates below: one global rate card |
| Models tracked | 4 |
Price by GPU
As of , per GPU-hour: Together AI’s on-demand rate for each card it lists, a cheaper market it also sells, and where the on-demand rate stands among the providers we track.
| GPU | Together AI on-demand, per GPU-hour | Against the market (median rate card) |
|---|---|---|
| NVIDIA B300 | $9.99 captured Deploy ↗ | 11% above the median $8.99 · 24 providers 15 of 24 list lower · lowest Deep Infra $5.89 · compare |
| NVIDIA B200 | $8.19 captured Deploy ↗ | 9% above the median $7.50 · 35 providers 23 of 35 list lower · lowest Beam $4.11 from · compare |
| NVIDIA H200 | $5.99 captured Deploy ↗ | 32% above the median $4.54 · 51 providers 37 of 51 list lower · lowest Beam $2.09 from · compare |
| NVIDIA H100 SXM | $3.99 captured Deploy ↗ | 13% above the median $3.52 · 38 providers 24 of 38 list lower · lowest Cudo Compute $1.79 GPU only |
The same, in words
- Together AI lists the NVIDIA B300 from $9.99 per GPU-hour on-demand. The on-demand rate is unchanged since first read, Sep 27, 2026 (7-day change: none). Against the market: 11% above the median rate card ($8.99 across 24 providers); 15 of the 24 providers we track list a lower on-demand rate, the lowest Deep Infra at $5.89 (Deep Infra’s rates, Together AI vs Deep Infra).
- Together AI lists the NVIDIA B200 from $8.19 per GPU-hour on-demand, $6.79 reserved (commit). The on-demand rate is unchanged since first read, Sep 7, 2026 (7-day change: none). Against the market: 9% above the median rate card ($7.50 across 35 providers); 23 of the 35 providers we track list a lower on-demand rate, the lowest Beam at $4.11 (Beam’s rates, Together AI vs Beam).
- Together AI lists the NVIDIA H200 from $5.99 per GPU-hour on-demand, $3.99 reserved (commit). The on-demand rate is unchanged since first read, Sep 7, 2026 (7-day change: none). Against the market: 32% above the median rate card ($4.54 across 51 providers); 37 of the 51 providers we track list a lower on-demand rate, the lowest Beam at $2.09 (Beam’s rates, Together AI vs Beam).
- Together AI lists the NVIDIA H100 SXM from $3.99 per GPU-hour on-demand, $3.19 reserved (commit). The on-demand rate is unchanged since first read, Sep 7, 2026 (7-day change: none). Against the market: 13% above the median rate card ($3.52 across 38 providers); 24 of the 38 providers we track list a lower on-demand rate, the lowest Cudo Compute at $1.79 (Cudo Compute’s rates).
All listed prices
One table per price type, never mixed. Every rate normalized to one GPU for one hour; a price opens the page it was read from, and 7d marks a rate that moved in the past seven days. The day each figure was first listed is in this page’s data, below.
On-demand
| GPU · what is sold | $ / GPU-hour | Read | Source |
|---|---|---|---|
| NVIDIA H100 SXM | $3.99 | captured | Deploy ↗ |
| NVIDIA H200 | $5.99 | captured | Deploy ↗ |
| NVIDIA B200 | $8.19 | captured | Deploy ↗ |
| NVIDIA B300 | $9.99 | captured | Deploy ↗ |
Reserved (committed use)
| GPU · what is sold | $ / GPU-hour | Read | Source |
|---|---|---|---|
| NVIDIA H100 SXMfrommin. commitment commit | $3.19 | captured | Reserve ↗ |
| NVIDIA H200frommin. commitment commit | $3.99 | captured | Reserve ↗ |
| NVIDIA B200frommin. commitment commit | $6.79 | captured | Reserve ↗ |
What else Together AI bills
The GPU-hour is not the whole bill. What Together AI's own pages say about the rest, read by hand and checked on ; each line links the page it was read from. "Not published" means we looked and the pages do not say. Every provider side by side.
| What | What Together AI states | Source |
|---|---|---|
| In the rate | Per-GPU hourly rate; node vCPU/RAM not priced separately; local NVMe per node (ephemeral); shared storage billed separately our reading | source checked |
| Data out | free: no data transfer fees in or out of the cluster | source checked |
| Storage | Shared filesystem $0.16/GiB-month; minimum volume size 1 TiB | source checked |
| While stopped | No pause: reserved billed for full term, on-demand stops only on termination; storage keeps billing without a cluster | source checked |
| Billing step | On-demand per hour of cluster runtime; reserved 1-90 days paid upfront, non-refundable; preemptible metered every 1-2 min | source checked |
Every rate on this page as data: /providers/together/prices.json — the same fields as prices.json, narrowed to this provider. CC BY 4.0, no key, no rate limit. On your own page: embed this table (/embed/providers/together, an iframe, no script).
Compare Together AI with
Every provider that lists at least two of the same GPUs on-demand, most shared GPUs first; each opens the side-by-side page.
- Amazon Web Services 4 shared GPUs
- Barrack AI 4 shared GPUs
- Crusoe 4 shared GPUs
- GPU.ai 4 shared GPUs
- Hyperstack 4 shared GPUs
- Lium 4 shared GPUs
- Modal 4 shared GPUs
- Nebius 4 shared GPUs
- Oracle Cloud (OCI) 4 shared GPUs
- RunPod 4 shared GPUs
- Verda 4 shared GPUs
- Bitdeer AI 3 shared GPUs
- Civo 3 shared GPUs
- CoreWeave 3 shared GPUs
- Daytona 3 shared GPUs
- Deep Infra 3 shared GPUs
- EmpirioLabs 3 shared GPUs
- fal 3 shared GPUs
- Fireworks AI 3 shared GPUs
- FPT AI Factory 3 shared GPUs
- Gcore 3 shared GPUs
- Google Cloud 3 shared GPUs
- HexGrid 3 shared GPUs
- HPC-AI.COM 3 shared GPUs
- Hyperbolic 3 shared GPUs
- Lyceum 3 shared GPUs
- Paperspace 3 shared GPUs
- Vast.ai 3 shared GPUs
- Atlas Cloud 2 shared GPUs
- Beam 2 shared GPUs
- Canopy Wave 2 shared GPUs
- Cerebrium 2 shared GPUs
- Dataoorts 2 shared GPUs
- E2E Networks 2 shared GPUs
- Elice Cloud 2 shared GPUs
- Geodd 2 shared GPUs
- GMI Cloud 2 shared GPUs
- Hugging Face 2 shared GPUs
- Jarvislabs 2 shared GPUs
- Koyeb 2 shared GPUs
- Lambda 2 shared GPUs
- Microsoft Azure 2 shared GPUs
- Neysa 2 shared GPUs
- Northflank 2 shared GPUs
- Scaleway 2 shared GPUs
- Seeweb 2 shared GPUs
- Sesterce 2 shared GPUs
- Taiga Cloud (Northern Data Group) 2 shared GPUs
Data updated: