All models

Llama 3.2 3B Instruct

MetaOpen weights131K contextTool calling

LMArena score 1109 · #360 overall · #360 in coding (2026-10-13)

Lowest price, from Lambda

Per million tokens

Input
$0.015
Output
$0.025
Cached input
—
Batch input / output
—
  • Up to 131K 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 Lambda: $0.20 a month.

Where to buy it (10)

Sorted by your monthly estimate. Prices per million tokens.

  • LambdaCheapest

    Input / output
    $0.015 / $0.025
    Context
    131K
    Per month
    $0.20

    lambda_ai/llama3.2-3b-instruct

  • DeepInfra

    Input / output
    $0.02 / $0.02
    Context
    131K
    Per month
    $0.24

    deepinfra/meta-llama/Llama-3.2-3B-Instruct

  • Novita AI

    Input / output
    $0.03 / $0.05
    Context
    33K
    Per month
    $0.40

    novita/meta-llama/llama-3.2-3b-instruct

  • Llamagate

    Input / output
    $0.04 / $0.08
    Context
    131K
    Per month
    $0.56

    llamagate/llama-3.2-3b

  • Together AI

    Input / output
    $0.06 / $0.06
    Context
    131K
    Per month
    $0.72

    together_ai/meta-llama/Llama-3.2-3B-Instruct

  • Cloudflare Workers AI

    Input / output
    $0.0509 / $0.335
    Context
    80K
    Per month
    $1.18

    cloudflare/@cf/meta/llama-3.2-3b-instruct

  • Hyperbolic

    Input / output
    $0.12 / $0.30
    Context
    33K
    Per month
    $1.80

    hyperbolic/meta-llama/Llama-3.2-3B-Instruct

  • Amazon Bedrock

    Input / output
    $0.15 / $0.15
    Context
    128K
    Per month
    $1.80

    meta.llama3-2-3b-instruct-v1:0

  • Vercel AI Gateway

    Input / output
    $0.15 / $0.15
    Context
    128K
    Per month
    $1.80

    vercel_ai_gateway/meta/llama-3.2-3b

  • IBM watsonx

    Input / output
    $0.15 / $0.15
    Context
    128K
    Per month
    $1.80

    watsonx/meta-llama/llama-3-2-3b-instruct

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

Similar models