Claude Fable 5$22.000/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-5.5$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/MGemini 1.5 Pro$2.375/MGemini 1.0 Ultra$12.000/MGemini 1.0 Pro$0.800/MClaude Fable 5$22.000/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-5.5$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/MGemini 1.5 Pro$2.375/MGemini 1.0 Ultra$12.000/MGemini 1.0 Pro$0.800/M
BETA
Subscribe
← Back to Index·All Together AI models
Together AI
Together AIFrontier

Mixtral 8x22B (Together)

Large MoE

Largest open-weights Mixtral MoE. Cheap-to-serve frontier-ish quality before Llama 3.1 405B and DeepSeek V3 took the open-weights lead.

Mixtral 8x22B (Together) is a frontier AI model from Together AI. It costs $1.200 per million input tokens and $1.200 per million output tokens (blended $1.200/M), with a 64K-token context window.

Profile inherited from upstream Mixtral 8x22B ↗ — this is a hosted variant of the same open-weights model.

Released Apr 2024Modalities textOfficial model page ↗Provider pricing ↗API docs ↗Compare with another model →Estimate monthly cost →
INPUT
$1.200/M
per million input tokens
OUTPUT
$1.200/M
per million output tokens
BLENDED 70/30
$1.200/M
default reference rate · how it's calculated →
CONTEXT
64K
64,000 tokens
What it's good at
  • Open-weights MoE
  • Cheap inference per active parameter
  • 64K context
Typical use cases
  • Self-hosted chat at scale
  • Fine-tune base
Benchmarks
vs. best public score
Scores inherited from Mixtral 8x22B — this is a hosted variant of the same open-weights model, so the underlying benchmark scores are identical.
MMLU78%
Multitask academic knowledge across 57 subjects.
GPQA Diamond36%
Graduate-level science questions, "Google-proof".
MATH42%
High-school competition math problems.
HumanEval76%
Python function synthesis from docstrings.
LMArena Elo1198 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 Together AI
See all 10 →
Llama 3.3 70B (Together)
Managed · 128K ctx
in $0.880/Mout $0.880/M
Llama 3.1 405B (Together)
Managed Large · 128K ctx
in $3.500/Mout $3.500/M
Qwen2.5-72B (Together)
Managed · 32K ctx
in $1.200/Mout $1.200/M
DeepSeek-R1 (Together)
Reasoning · 64K ctx
in $3.000/Mout $7.000/M
DBRX Instruct (Together)
Enterprise · 32K ctx
in $1.200/Mout $1.200/M
Falcon 40B (Together)
Open · 2K ctx
in $0.900/Mout $0.900/M
Frequently asked questions

How much does Mixtral 8x22B (Together) cost?

Mixtral 8x22B (Together) costs $1.200 per million input tokens and $1.200 per million output tokens, for a blended reference rate of $1.200 per million tokens.

What is Mixtral 8x22B (Together)'s context window?

Mixtral 8x22B (Together) supports up to 64K tokens of context (64,000 tokens).

What is Mixtral 8x22B (Together) best for?

Mixtral 8x22B (Together) is well suited to Open-weights MoE, Cheap inference per active parameter and 64K context.

Who makes Mixtral 8x22B (Together)?

Mixtral 8x22B (Together) is developed and served by Together AI. It was released in Apr 2024.