Inference cloud
How Fireworks AI compares
unverifiedcatalog v1
Serverless per-token rates and on-demand GPU-hour rates.
Regions and compliance
Regions: not published / unverified
Compliance: not published / unverified
Other published facts
- Batch share of serverless: 0.5docs.fireworks.aichecked 2026-10-20
Token offerings
GPT OSS 120B standard
serverless_api · gpt-oss-120b · open-weight true
- Input / 1M
- 0.15docs.fireworks.aichecked 2026-10-20“OpenAI GPT OSS 120B | $0.15 / $0.015 / $0.60”
- Cached input / 1M
- 0.015docs.fireworks.aichecked 2026-10-20“OpenAI GPT OSS 120B | cached input | $0.015”
- Output / 1M
- 0.6docs.fireworks.aichecked 2026-10-20“OpenAI GPT OSS 120B | output | $0.60”
Context not published · tokens/s not published · TTFT not published
Standard cell is input / cached input / output. Open-weight flag is cited from GroqDocs.
DeepSeek V4.1 Flash standard
serverless_api · deepseek-v4.1-flash · open-weight not published / unverified
- Input / 1M
- 0.3docs.fireworks.aichecked 2026-10-20“DeepSeek V4.1 Flash | $0.30 / $0.006 / $1.20”
- Cached input / 1M
- 0.006docs.fireworks.aichecked 2026-10-20“DeepSeek V4.1 Flash | cached input | $0.006”
- Output / 1M
- 1.2docs.fireworks.aichecked 2026-10-20“DeepSeek V4.1 Flash | output | $1.20”
Context not published · tokens/s not published · TTFT not published
Compute SKUs
H100 80 GB on-demand 1× H100 · gpu hour · in the run-cost model
On demand
8fireworks.aichecked 2026-10-20“H100 80 GB GPU | $0.134 per minute | $8.00 per hour”Reserved
not published / unverifiedfireworks.aichecked 2026-10-20“No reserved rate was listed.”Spot
not published / unverifiedfireworks.aichecked 2026-10-20“No spot rate was listed.”Page says pay per GPU second. The hourly column is stored. Billing in Compare is scale-to-zero (busy seconds), using $8/hour.
B200 180 GB on-demand 1× B200 · gpu hour · shown raw, not in C_1k
On demand
13fireworks.aichecked 2026-10-20“B200 180 GB GPU | $0.217 per minute | $13.00 per hour”Reserved
not published / unverifiedfireworks.aichecked 2026-10-20“No reserved rate was listed.”Spot
not published / unverifiedfireworks.aichecked 2026-10-20“No spot rate was listed.”