Enterprise reserved instance
Priced on request
Reserved GPUs; multi-region deployments, custom optimizations, better performance, and BYOC compatibility
Checked 7 Oct 2026
Plans and prices
Fireworks AI pricing
Managed training
Not listed
LoRA SFT, LoRA DPO, Full Param SFT, and Full Param DPO; rates range by model size
Checked 7 Oct 2026
On-demand deployments
Not listed
H100 and H200: $0.134/minute or $8.00/hour; B200: $0.217/minute or $13.00/hour; B300: $0.250/minute or $15.00/hour; GB300: $0.334/minute or $20.00/hour
Checked 7 Oct 2026
Serverless inference — embeddings
Not listed
Prices vary by embedding model parameter count
Checked 7 Oct 2026