Claude Fable 5$22.000/MClaude Opus 5$11.000/MClaude Opus 4.8$11.000/MClaude Opus 4.7$11.000/MClaude Opus 4.6$11.000/MClaude Opus 4.5$33.000/MClaude Sonnet 3.7$6.600/MClaude Opus 3$33.000/MClaude 2.1$12.800/MClaude 2$12.800/MGPT-5.6 Sol$12.500/MGPT-5.6 Terra$5.000/MGPT-5.5$12.500/MGPT-5.2$5.425/MGPT-5.2-Codex$5.425/MGPT-5$3.875/MGPT-4.5$97.500/MGPT-4 Turbo Preview$16.000/MGPT-4$39.000/MGPT-4-32k$78.000/Mo3$19.000/Mo3-mini$2.090/Mo4-mini$2.090/Mo1$28.500/Mo1-mini$5.700/Mo1-preview$28.500/MGemini 3.5 Pro$5.000/MGemini 3.1 Pro$5.000/MGemini 3 Pro$5.000/MGemini 2.5 Pro$3.875/MClaude Fable 5$22.000/MClaude Opus 5$11.000/MClaude Opus 4.8$11.000/MClaude Opus 4.7$11.000/MClaude Opus 4.6$11.000/MClaude Opus 4.5$33.000/MClaude Sonnet 3.7$6.600/MClaude Opus 3$33.000/MClaude 2.1$12.800/MClaude 2$12.800/MGPT-5.6 Sol$12.500/MGPT-5.6 Terra$5.000/MGPT-5.5$12.500/MGPT-5.2$5.425/MGPT-5.2-Codex$5.425/MGPT-5$3.875/MGPT-4.5$97.500/MGPT-4 Turbo Preview$16.000/MGPT-4$39.000/MGPT-4-32k$78.000/Mo3$19.000/Mo3-mini$2.090/Mo4-mini$2.090/Mo1$28.500/Mo1-mini$5.700/Mo1-preview$28.500/MGemini 3.5 Pro$5.000/MGemini 3.1 Pro$5.000/MGemini 3 Pro$5.000/MGemini 2.5 Pro$3.875/M
Fireworks AI
Fireworks AI
Frontier

Llama 3.1 405B (fw)

Serverless

Largest dense Llama. Used as a quality benchmark for open weights, heavy to run, often hosted via Together / DeepInfra / Cerebras.

Llama 3.1 405B (fw) is a frontier AI model from Fireworks AI. It costs $3.000 per million input tokens and $3.000 per million output tokens (blended $3.000/M), with a 131,000-token context window.

Profile inherited from upstream Llama 3.1 405B — this is a hosted variant of the same open-weights model.

INPUT
$3.000/M
per million input tokens
OUTPUT
$3.000/M
per million output tokens
BLENDED 70/30
$3.000/M
unchanged since 3 May
CONTEXT
131,000
tokens
What it is good at
  • Top open-weights quality
  • Permissive license
  • Strong reasoning
Typical use cases
  • Open-weights quality ceiling
  • Synthetic data generation
  • Distillation source

Benchmarks

vs. best public score
Scores inherited from Llama 3.1 405B — this is a hosted variant of the same open-weights model, so the underlying benchmark scores are identical.
MMLU88%
Multitask academic knowledge across 57 subjects.
Graduate-level science questions, "Google-proof".
MATH73%
High-school competition math problems.
Python function synthesis from docstrings.
LMArena Elo1290 Elo
Crowd-sourced head-to-head preference Elo rating.
Hand-curated from each provider's published reports and public leaderboards. Methodology varies across sources, treat as directional rather than authoritative.

More from Fireworks AI

See all 7

Frequently asked questions

How much does Llama 3.1 405B (fw) cost?

Llama 3.1 405B (fw) costs $3.000 per million input tokens and $3.000 per million output tokens, for a blended reference rate of $3.000 per million tokens.

What is Llama 3.1 405B (fw)'s context window?

Llama 3.1 405B (fw) supports up to 131,000 tokens of context in a single request.

What is Llama 3.1 405B (fw) best for?

Llama 3.1 405B (fw) is well suited to Top open-weights quality, Permissive license and Strong reasoning.

Who makes Llama 3.1 405B (fw)?

Llama 3.1 405B (fw) is developed and served by Fireworks AI.

Terms used on this page