Hosted open + custom is a Hosted from Fireworks AI (United States).
One-line fit: best used for Production inference.
Key specifications
| Provider | Fireworks AI |
|---|---|
| Model | Hosted open + custom |
| Type | Hosted |
| Context window | Per model |
| Pricing (indicative) | API |
| Company category | Inference cloud |
| Region | United States |
| Official site | https://fireworks.ai |
Benchmarks and public standing
Speed/cost product metrics
Treat scores as directional. SWE-bench, Terminal-Bench, LMSYS Arena, Artificial Analysis, and vendor cards use different harnesses and are not always comparable 1:1. Re-check the latest official model card before decisions.
What this model is good at
Production inference
This aligns with Fireworks AI’s broader strengths:
- Low-latency open inference
- Production serving
- LoRA deploy
Limitations and watch-outs
Not base trainer of frontier closed
- Rate limits, regional availability, and data retention policies vary by plan.
- Agent harness quality (Cursor, Claude Code, Codex, custom tools) can change outcomes more than raw model Elo.
- Open-weight availability (if any) is separate from hosted API quality and safety filters.
Ideal users
- Product engineers shipping features that match: Production inference
- Teams standardizing on the Fireworks AI ecosystem
- Agent builders who need a Hosted
When to pick something else
- Cheaper volume: compare lower tiers from the same lab or open Chinese/EU alternatives.
- Maximum hard SWE thoughtfulness: compare Claude Fable-class and other coding flagships.
- Giant multimodal corpora / Workspace: compare Gemini-class models.
- Self-host / open weights: Llama, Qwen, DeepSeek, GLM, Mistral open lines.
Parent company
Fireworks AI – High-performance inference platform for open and custom models.
Related models from Fireworks AI
Data is curated for AIForumSphere model directory mid-2026 directional figures.
