Together AI alternatives, like-for-like
An alternative is only listed where it offers the same GPU model, the same form and the same tier as a Together AI offer. Cheaper does not mean equal: check reliability, region and data-out fees.
B200 SXM · Secure
Together AI: $8.19 per GPU-hr.
| Provider | $ / GPU-hr | vs Together AI | Notes |
|---|---|---|---|
| Vultr | from $3.50 | −57% | starting from; 8-GPU node |
| Hyperstack | $6.00 | −27% | |
| Lambda | $7.69 | −6% | 8x HGX B200; announced price effective 2026-11-16 |
| RunPod | $7.99 | −2% | |
| Nebius | $8.50 | +4% | |
| CoreWeave | $8.60 | +5% | per GPU-hour from the 8-GPU node price ($68.80) |
| Verda (ex-DataCrunch) | $9.138 | +12% |
H100 SXM · Secure
Together AI: $3.99 per GPU-hr.
| Provider | $ / GPU-hr | vs Together AI | Notes |
|---|---|---|---|
| Voltage Park | from $1.99 | −50% | FAQ: "starting at $1.99 an hour without a contract" |
| Vultr | from $2.30 | −42% | HGX H100, starting from; 8-GPU node |
| Hyperstack | $3.20 | −20% | |
| Jarvis Labs | $3.49 | −13% | |
| Crusoe | $3.90 | −2% | HGX H100; new price $4.50 from Dec 1, 2026 |
| RunPod | $3.99 | +0% | |
| DigitalOcean GPU Droplets | $4.41 | +11% | |
| Nebius | $4.50 | +13% | |
| Lambda | $4.59 | +15% | 8x HGX H100 instance; announced price effective 2026-11-16 |
| Verda (ex-DataCrunch) | $4.987 | +25% | |
| CoreWeave | $6.16 | +54% | per GPU-hour from the 8-GPU node price ($49.24) |
| AWS | $6.88 | +72% | p5.48xlarge/8, us-east-1 |
| Oracle Cloud (OCI) | $10.00 | +151% | BM.GPU.H100.8, flat per-GPU rate |
| Google Cloud | $11.06 | +177% | a3-highgpu-8g/8, us-central1 |
| Microsoft Azure | $12.29 | +208% | ND96isr H100 v5/8, East US |
H100 SXM · Spot
Together AI: $1.99 per GPU-hr.
| Provider | $ / GPU-hr | vs Together AI | Notes |
|---|---|---|---|
| Nebius | from $0.79 | −60% | preemptible, "from" |
| Jarvis Labs | $1.54 | −23% | |
| Verda (ex-DataCrunch) | $2.494 | +25% |
H200 SXM · Secure
Together AI: $5.99 per GPU-hr.
| Provider | $ / GPU-hr | vs Together AI | Notes |
|---|---|---|---|
| Hyperstack | $3.99 | −33% | |
| Crusoe | $4.29 | −28% | HGX H200; new price $5.40 from Dec 1, 2026 |
| DigitalOcean GPU Droplets | $4.47 | −25% | |
| Jarvis Labs | $4.59 | −23% | |
| RunPod | $5.29 | −12% | |
| Nebius | $5.40 | −10% | |
| CoreWeave | $6.31 | +5% | per GPU-hour from the 8-GPU node price ($50.44) |
| Verda (ex-DataCrunch) | $6.44 | +8% | |
| AWS | $7.91 | +32% | p5en.48xlarge/8, us-east-1 |
| Microsoft Azure | $10.60 | +77% | ND96isr H200 v5/8, East US 2 |
H200 SXM · Spot
Together AI: $2.99 per GPU-hr.
| Provider | $ / GPU-hr | vs Together AI | Notes |
|---|---|---|---|
| Nebius | from $0.79 | −74% | preemptible, "from" |
| Jarvis Labs | $2.29 | −23% | |
| CoreWeave | $2.62 | −12% | spot per GPU-hour from the 8-GPU node price ($20.93) |
| Verda (ex-DataCrunch) | $3.22 | +8% |
Sponsored row: none yet. Advertise. Prices from vendor pages 2026-11-10.