stacksherpa

API provider directory

Fireworks AI

Fast inference platform with broad model selection. Hosts GLM-4.7, Qwen3 (8B/30B), Kimi K2.5, and many open models. FireFunction models for reliable tool/function calling. Compound AI system support. Cached input tokens at 50% off, batch at 50% off, no premium for fine-tuned model inference. OpenAI-compatible API. Pricing: Qwen3 8B ~$0.20/1M, Qwen3 30B ~$0.26/1M, GLM-4.7 ~$0.60/$2.20 per 1M tokens.

website | docs | pricing page | github | npm: fireworks-js

Overview

CategoryAi Image
ComplianceSOC2, GDPR
Self-HostableNo
On-PremNo
Best Forstartup, growth, enterprise
Last Verified2026-02-12

Strengths & Weaknesses

Strengths:Weaknesses:

When to Use

Best when:Avoid if:

Alternatives

together-ai, groq, replicate

Fireworks AI - Ai Image - stacksherpa