Fireworks adds two new serverless models, including its cheapest input rate yet
Fireworks added Muse Glimmer 30B and NVIDIA Nemotron 3.5 Lightning 30B A3B to its serverless rate card; Nemotron 3.5 Lightning's $0.05 input rate is now the cheapest published on the entire card.
Cheapest published serverless input rate was OpenAI GPT OSS 20B at $0.07 per 1M tokens; the model catalog did not include Muse Glimmer or any Nemotron 3.5 variant.
NVIDIA Nemotron 3.5 Lightning 30B A3B is priced at $0.05 / $0.01 / $0.20 per 1M tokens (Standard-only), the new cheapest input rate on the card. Muse Glimmer 30B is priced at $0.35 / $0.04 / $1.50 Standard and $0.525 / $0.06 / $2.25 Priority.
Fireworks’ serverless pricing docs picked up two new priced model rows in this capture, with no other changes elsewhere on the pricing or docs surfaces.
Muse Glimmer 30B joins the card at $0.35 / $0.04 / $1.50 per 1M tokens (input / cached input / output) on the Standard path, with a Priority path at $0.525 / $0.06 / $2.25 — a flat 1.5x uplift consistent with several other models on the card.
NVIDIA Nemotron 3.5 Lightning 30B A3B is the more notable addition: priced at $0.05 / $0.01 / $0.20 per 1M tokens on Standard only (no Priority path published), its $0.05 input rate undercuts the prior cheapest published rate on the card, OpenAI GPT OSS 20B at $0.07. This continues Fireworks’ pattern of expanding its open-weight model catalog with granular, size-appropriate pricing rather than a flat rate across all hosted models.
Fireworks quietly added two new rows to its serverless per-model rate card: Muse Glimmer 30B ($0.35 / $0.04 / $1.50 Standard, $0.525 / $0.06 / $2.25 Priority — a flat 1.5x uplift) and NVIDIA Nemotron 3.5 Lightning 30B A3B ($0.05 / $0.01 / $0.20, Standard-only). Nemotron 3.5 Lightning's $0.05 input rate is now the cheapest published on the entire serverless card, undercutting the previous low of OpenAI GPT OSS 20B at $0.07. No other pricing page or docs surface changed in this capture.