Fireworks AI
Fast, pay-per-token inference cloud for open-source and fine-tuned LLMs, with dedicated GPU deployments and enterprise compliance.
Last verified:
Fireworks AI is a San Mateo-based inference platform offering per-token serverless pricing (Standard/Priority/Fast speed tiers), on-demand GPU deployments (H100 through B300), and managed fine-tuning/training across open-source models. It provides a REST API, Python SDK, firectl CLI, and a documentation-search MCP server for coding agents, alongside SOC 2 Type II, HIPAA, and GDPR compliance with Google/OIDC/SAML SSO. Founded in 2022 by ex-Meta PyTorch lead Lin Qiao, Fireworks raised a $1.505B Series D in July 2026 at a $17.5B valuation, reporting over $1B ARR and 40+ trillion tokens served daily.
Pricing
| Tier | Price | Key features |
|---|---|---|
| Serverless Inference | Custom | Standard, Priority, and Fast speed tiers, Embeddings: $0.008-$0.10 per 1M input tokens, High rate limits, no long-term commitment |
| On-Demand Deployments | $7/usage | H100/H200 80GB: $7.00/hr, B200 180GB: $10.00/hr, B300 288GB: $12.00/hr, No extra charge for startup time |
| Fine-Tuning / Training | $0.5/usage | LoRA SFT up to 16B: $0.50-$2.00 per 1M tokens, 80B-300B full-param DPO: up to $24.00 per 1M tokens, >300B models: up to $40.00 per 1M tokens |
| Enterprise | Custom | Faster speeds, lower costs, higher rate limits, Dedicated deployments and data residency support, SOC2/HIPAA/GDPR compliance bundle, SSO |
Source: https://fireworks.ai/pricing
Agent-readiness
- Discoverability
- 70/100
- Machine-readable content
- 82/100
- Programmatic access
- 90/100
- MCP / agent support
- 55/100
- Self-serve transactability
- 85/100
- Transparency
- 82/100
Security & compliance
- SSO / SAML
- Yes
- Encryption at rest
- Not publicly documented
- Encryption in transit
- Yes
- Audit log
- Not publicly documented
- Pentest report
- Not publicly documented
"Not publicly documented" means Crail found no public evidence either way — not a confirmed absence. See our methodology. Trust center: https://trust.fireworks.ai.
FAQ
Does Fireworks AI have a free tier?
Yes, Fireworks AI offers a free tier or free trial.
Can you buy Fireworks AI without talking to sales?
Yes, Fireworks AI can be purchased self-serve without a sales call.
Does Fireworks AI have an official MCP (Model Context Protocol) server?
Yes, Fireworks AI publishes an official MCP server.
What compliance certifications does Fireworks AI hold?
Fireworks AI holds: SOC 2 Type II, HIPAA, GDPR.
Reviews
No reviews yet for Fireworks AI. Be the first to submit a verified review.
Methodology & disclosure
Data verified via agent crawled, last checked 2026-07-27. Agent-readiness methodology version crail-ar-v0.1. No vendor payment influences these scores — see our methodology page.