{T}
TokenMath.net
Updated 2026-09-22 11:16 UTC
Head-to-Head Specification Battle

Llama 4 Scout (17B Active / 109B MoE) — Together AI vs Command R (08-2024)

Side-by-side technical and economic comparison between Meta / Together's Llama 4 Scout (17B Active / 109B MoE) — Together AI and Cohere's Command R (08-2024).

Endpoint Availability & Transition Notice

Llama 4 Scout (17B Active / 109B MoE) — Together AI: Classified as a Legacy Reference model (Deprecated on Together AI (Reference)).

Command R (08-2024): Classified as a Legacy Reference model (Tracked Production / Legacy Family).

Core Technical & Pricing Comparison
MetricLlama 4 Scout (17B Active / 109B MoE) — Together AICommand R (08-2024)Advantage
Provider OrganizationMeta / TogetherCohere
Standard Input / 1M$0.18$0.15Command R (08-2024) (17% lower)
Cached Input / 1MNone$0.015Command R (08-2024) lower
Output / 1M$0.59$0.60Llama 4 Scout (17B Active / 109B MoE) — Together AI lower
Context Window328k128kLlama 4 Scout (17B Active / 109B MoE) — Together AI (328k)
Max Generation Tokens16.4k4.1kLlama 4 Scout (17B Active / 109B MoE) — Together AI
Tokenizer FamilyMeta Llama 3/4 Tiktoken (128k vocabulary)Cohere BPE (~256k vocabulary)

Interactive Side-by-Side Cost Simulator

Live Delta
Input Tokens / Req2,500
Output Tokens / Req800
Daily Requests10,000
Llama 4 Scout (17B Active / 109B MoE) — Together AI
$276.60 / month
$0.00092 / request
Command R (08-2024)LOWER COST
$205.88 / month
$0.00069 / request
Selecting Command R (08-2024) saves $70.725 / month (25.6% savings) over Llama 4 Scout (17B Active / 109B MoE) — Together AI.

When to Choose Llama 4 Scout (17B Active / 109B MoE) — Together AI

Opt for Llama 4 Scout (17B Active / 109B MoE) — Together AI when your engineering requirements prioritize Meta / Together's ecosystem, specific tokenizer efficiencies (Calibrated Llama 3 Tokenizer (±3%)), or when your expected prompt-to-completion ratios favor its $0.18/M input rate.

When to Choose Command R (08-2024)

Opt for Command R (08-2024) when looking for Cohere's tooling integration, specific context window depth (128k tokens), or when output generation volume favors its $0.6/M rate.