Side by side pricing and context for Llama 3.3 70B (Groq) and Llama 3.3 70B (Fireworks). Rates are curated catalog values with last verified dates on each model page.
Llama 3.3 70B (Groq)
Meta Llama 3.3 70B served on Groq for speed-sensitive chat. Price is for Groq hosting. Compare latency needs vs OpenAI/Anthropic quality.
Input $0.590 / Output $0.790 per 1M
Context 128K · Approx
Llama 3.3 70B (Fireworks)
Llama 3.3 70B on Fireworks for fast open-weight inference. Useful when comparing Fireworks vs Together vs Groq hosting costs.
Input $0.900 / Output $0.900 per 1M
Context 128K · Approx
Quick take
- Llama 3.3 70B (Groq) is cheaper on input in this catalog snapshot.
- Llama 3.3 70B (Groq) is cheaper on output in this catalog snapshot.
- Both last verified on their model pages (2026-08-02 / 2026-08-02).
Rate table
| Metric | Llama 3.3 70B (Groq) | Llama 3.3 70B (Fireworks) |
|---|---|---|
| Input / 1M | $0.590 | $0.900 |
| Output / 1M | $0.790 | $0.900 |
| Cached input / 1M | n/a | n/a |
| Context | 128K | 128K |
| Tokenizer | Approx | Approx |