How is Fireworks serverless priced?

USD = (uncached input ÷ 1M × input $/1M) + (cached ÷ 1M × cached $/1M) + (output ÷ 1M × output $/1M). Example: DeepSeek V4.1 Flash Standard at $0.30/$0.006/$1.20 with 1M in + 0 cached + 200k out ≈ $0.54.

What is Priority vs Standard?

Each headline model lists Standard and Priority rates (input / cached / output). Some Fast variants have no Priority column — this calculator falls back to Standard rates and shows a notice.

Optional Batch applies 50% of Standard serverless rates on input, cached input, and output per Fireworks docs. Confirm Batch eligibility on the official docs.

Are size-tier and embeddings included?

No. “Other base models by size” and embeddings are omitted in v1 (footnote only). Estimates only; docs.fireworks.ai/serverless/pricing prevails.

Is any data uploaded?

No. These calculators and QR codes run in your browser.

Calculators

Fireworks Token Cost Calculator

Estimate Fireworks.ai serverless inference cost from model, input/cached/output tokens, and Standard or Priority tier. Headline presets include DeepSeek V4.1 Flash ($0.30/$0.006/$1.20), GLM 5.3 ($1.40/$0.26/$4.40), Kimi K3 ($3/$0.30/$15), and GPT-OSS 120B ($0.15/$0.015/$0.60) per 1M tokens. Optional Batch is 50% of Standard. Size-tier “Other base models” and embeddings are out of scope for v1. Estimates only; official Fireworks docs prevail.

Math runs in your browser. Nothing is uploaded.

Model

Service tier

Priority is not available for this model — Standard rates are used.

Input tokens

Of which cached input tokens

Output tokens

Input price ($ / 1M tokens)

Cached input ($/1M)

Output price ($ / 1M tokens)

Batch half-price (50% of Standard)

Input cost
—

Cached input cost
—

Output cost
—

Estimated USD cost
—

Monthly USD
—

Estimates only; official vendor pricing prevails (2026-10-30). Cost = uncached in÷1M×rate + cached÷1M×rate + out÷1M×rate. Examples (Standard): DeepSeek V4.1 Flash $0.30/$0.006/$1.20; GLM 5.3 $1.40/$0.26/$4.40; Kimi K3 $3/$0.30/$15; GPT-OSS 120B $0.15/$0.015/$0.60. Priority has its own column; Batch = 50% of Standard. Size-tier and embeddings omitted.Official pricing: Fireworks serverless

Groq Token Cost Calculator · Mistral Token Cost Calculator · DeepSeek Token Cost Calculator · LLM API Cost Calculator · Together AI Token Cost Calculator

How it works

How is Fireworks serverless priced?: USD = (uncached input ÷ 1M × input $/1M) + (cached ÷ 1M × cached $/1M) + (output ÷ 1M × output $/1M). Example: DeepSeek V4.1 Flash Standard at $0.30/$0.006/$1.20 with 1M in + 0 cached + 200k out ≈ $0.54.
What is Priority vs Standard?: Each headline model lists Standard and Priority rates (input / cached / output). Some Fast variants have no Priority column — this calculator falls back to Standard rates and shows a notice.
How does Batch work?: Optional Batch applies 50% of Standard serverless rates on input, cached input, and output per Fireworks docs. Confirm Batch eligibility on the official docs.
Are size-tier and embeddings included?: No. “Other base models by size” and embeddings are omitted in v1 (footnote only). Estimates only; docs.fireworks.ai/serverless/pricing prevails.
Is any data uploaded?: No. These calculators and QR codes run in your browser.

Fireworks Token Cost Calculator

More in this group

How it works