Appndo
All models

Llama3 70B Instruct

MetaOpen weights8K context

LMArena score 1221 · #309 overall · #312 in coding (2026-10-13)

Lowest price, from Novita AI

Per million tokens

Input
$0.51
Output
$0.74
Cached input
—
Batch input / output
—
  • Up to 8K output tokens per request.

Estimate your monthly cost

Est. per month = input tokens × input price + output tokens × output price. Prices are per million tokens, so 10M input tokens at $3.00 cost $30.

From Novita AI: $6.58 a month.

Where to buy it (4)

Sorted by your monthly estimate. Prices per million tokens.

  • Novita AICheapest

    Input / output
    $0.51 / $0.74
    Context
    8K
    Per month
    $6.58

    novita/meta-llama/llama-3-70b-instruct

  • Vercel AI Gateway

    Input / output
    $0.59 / $0.79
    Context
    8K
    Per month
    $7.48

    vercel_ai_gateway/meta/llama-3-70b

  • Replicate

    Input / output
    $0.65 / $2.75
    Context
    8K
    Per month
    $12.00

    replicate/meta/llama-3-70b

  • Replicate

    Input / output
    $0.65 / $2.75
    Context
    8K
    Per month
    $12.00

    replicate/meta/llama-3-70b-instruct

  • Amazon Bedrock

    Input / output
    $2.65 / $3.50
    Context
    8K
    Per month
    $33.50

    meta.llama3-70b-instruct-v1:0

Cloud and hosted sellers may add regional pricing, minimums or platform fees. Supports: —.

Similar models