Together AI
On-demand dedicated endpoints billed per minute: A100 40GB $2.40/hr, A100 80GB $2.59/hr, H100 80GB $3.36/hr (up to ~43% off legacy). Also serverless token APIs and clusters (H100 f
Together AI scores 4.6/10 for free value in Functions. The main constraint: Paid — H100 80GB $3.36/hr; A100 80GB $2.59/hr; A100 40GB $2.40/hr. On-demand dedicated endpoints billed per minute: A100 40GB $2.40/hr, A100 80GB $2.59/hr, H100 80GB $3.36/hr (up to ~43% off legacy). Also serverless token APIs and clusters (H100 from $5.49/hr).
Free value score
4.6 / 10Show the math
Key metrics
What you get
Paid — H100 80GB $3.36/hr; A100 80GB $2.59/hr; A100 40GB $2.40/hr. On-demand dedicated endpoints billed per minute: A100 40GB $2.40/hr, A100 80GB $2.59/hr, H100 80GB $3.36/hr (up to ~43% off legacy). Also serverless token APIs and clusters (H100 from $5.49/hr).
A100 80GB at $2.59/hr and H100 at $3.36/hr for dedicated endpoints is competitive for guaranteed-capacity inference with no queue.
Gotchas
No gotchas documented yet. Contribute one if you know a constraint we missed.
Sources
-
"A100 SXM 40GB, $2.40/hour, -20% ; A100 SXM 80GB, $2.59/hour, -20% ; H100 80GB, $3.36/hour, -43%"
https://www.together.ai/blog/on-demand-dedicated-endpoints -
"NVIDIA HGX H100 $5.49 ... NVIDIA HGX H200 $6.79"
https://www.together.ai/pricing
Alternatives
Same category · similar quota$5/mo unlocks Workers + Containers + Durable Objects + R2 ($0.015/GB-month standard, Class A $4.50/M, Class B $0.36/M) with no egress fees.