Fireworks AI — pricing, limits & comparison
Generated 2026-11-03 12:14 UTC · JSON
Fireworks AI scores 74.29 for unified access (one api, many models) (LLM gateways & inference providers). Agent-friction 75/100.
Pricing
| Plan | Base / mo | Included | Overage |
|---|---|---|---|
| gpt-oss-120B (serverless, per-token) | $0 | — | $0.15/1 input_mtok, $0.60/1 output_mtok |
Cost at scale — LLM gateways & inference providers
| gpt-oss-120B (10M in / 2M out) |
|---|
| $2.70 |
Agent-friction breakdown — 75/100
| Check | Points | Pass |
|---|---|---|
| Test credentials without a credit card | 20 | ✓ |
| Sandbox or test mode | 15 | × |
| Setup with no human-only steps | 15 | ✓ |
| Official CLI | 10 | ✓ |
| Official MCP server | 10 | × |
| Markdown docs, current with the SDK | 10 | ✓ |
| Typed SDK for main languages | 10 | ✓ |
| Error messages that name the fix | 10 | ✓ |
Official SDKs
Python (fireworks-ai) · TypeScript (fireworks-ai)
Health & reliability — 100/100
| Incidents (90 days) | 0 · status page |
| Changelog entries (90 days) | 1 |
| Deprecations (90 days) | 0 |
| Downloads trend | +31.9% (fireworks-ai on pypi) |
| Last release | 2026-10-21 |
| Ownership events (90 days) | 0 |
Sources: pypi.org · status.fireworks.ai. Captured 2026-11-03.