Fireworks Announces Major Price Decrease on H100 GPU, Now Just $4.00/hour (31% Off)

Prompt caching discounts now available for enterprise deployments

FLUX.1 Kontext Max image model added at $0.08 per image

OpenAI gpt OSS 20b model added at $0.07 input, $0.30 output per 1M tokens

B200 180 GB GPU price decreased from $11.99/hour to $9.00/hour (25% decrease)

OpenAI gpt OSS 120b model added at $0.15 input, $0.60 output per 1M tokens

FLUX.1 Kontext Pro image model added at $0.04 per image

Batch inference pricing added at 50% discount of serverless pricing for input and output tokens

VLM supervised fine-tuning (fine-tuning with images) added, billed per token

Streaming ASR v2 model added at $0.0035 per audio minute

Kimi K2 Instruct model added at $0.60 input, $2.50 output per 1M tokens

Qwen3 Coder 480B model added at $0.45 input, $1.80 output per 1M tokens

Fine tuning pricing expanded from 3 tiers to 4 tiers: split DeepSeek R1/V3 ($10.00) into Models 80B-300B ($6.00) and Models >300B ($10.00)

H200 141 GB GPU price decreased from $6.99/hour to $6.00/hour (14% decrease)

H100 80 GB GPU price decreased from $5.80/hour to $4.00/hour (31% decrease)