Fireworks AI pricing: serverless per-token rates
Checked Nov 10, 2026 on Fireworks AI’s official pages (sources below). Prices change often: confirm with the provider before you buy.
Fireworks AI hosts open models behind a serverless API billed per token, with dedicated GPU deployments also available.
Figures below are quoted from fireworks.ai/pricing as read on Nov 10, 2026. Prices change: confirm on Fireworks AI's page before you buy.
Serverless, Standard tier (per 1M tokens)
- NVIDIA Nemotron 3.5 Lightning 30B A3B
- Input$0.05
- Cached input$0.01
- Output$0.20
- OpenAI GPT OSS 120B
- Input$0.15
- Cached input$0.015
- Output$0.60
- GLM 5.3 Flash
- Input$0.15
- Cached input$0.03
- Output$0.50
- DeepSeek V4.1 Flash
- Input$0.30
- Cached input$0.006
- Output$1.20
- MiniMax M3
- Input$0.30
- Cached input$0.06
- Output$1.20
- Nemotron 3 Ultra
- Input$0.60
- Cached input$0.12
- Output$2.40
- GLM 5.3
- Input$1.40
- Cached input$0.26
- Output$4.40
- Qwen 3.8 Max
- Input$2.00
- Cached input$0.25
- Output$6.00
- Kimi K3
- Input$3.00
- Cached input$0.30
- Output$15.00
Batch is 50% off and Spot is 30% off these Standard rates. Priority costs more; for example Kimi K3 Priority is $3.75 input, $0.375 cached and $18.75 output.
Fireworks AI pricing FAQ
What is the cheapest model on Fireworks AI?
Of the rates listed, Nemotron 3.5 Lightning 30B A3B was the lowest at $0.05 input and $0.20 output per 1M tokens (Standard tier).
Where do these Fireworks AI prices come from?
Fireworks AI's own pricing page (https://fireworks.ai/pricing), read on Nov 10, 2026. We do not estimate bills or add up costs across providers.
Sources (official pages, read Nov 10, 2026)
Spot something out of date? Email hello@youraitools.co and we’ll fix it.