Stable LM 3B on Replicate API

Name: Stable LM 3B on Replicate API
Brand: Stability AI
Price: 0.05 USD

StableLM · Stability AI

Serverless

Last refreshed 2026-04-15. Next refresh: weekly.

Why use Stable LM 3B on Replicate API?

Replicate API offers Stable LM 3B with pay-as-you-go pricing at $0.05/1M input tokens. Replicate is a cloud-based platform that enables users to run machine learning models easily and efficiently.

Input / 1M

$0.050

Output / 1M

$0.25

Cache

Not sourced

Batch

Not sourced

Setup recipe

Python + curl

Install

pip install replicate

Auth

export REPLICATE_API_TOKEN=...

Call

import replicate
output = replicate.run(
    "stable-lm-3b",
    input={"prompt": "Hello"}

Model ID

stable-lm-3b

Request example

import replicate

# reads REPLICATE_API_TOKEN from env
# stable-lm-3b format: "owner/model-name" (latest version) or "owner/model-name:version-hash"
output = replicate.run(
    "stable-lm-3b",
    input={"prompt": "Hello"}
)
# Output is a list or generator depending on the model
print("".join(output))

Gotchas

Replicate uses "owner/model-name" format (e.g. "meta/meta-llama-3-8b-instruct") for the latest version, or "owner/model-name:version-sha" to pin to a specific version. The REST endpoint splits owner and model-name into the path: /v1/models/{owner}/{model-name}/predictions.
The examples expect REPLICATE_API_TOKEN; rename it only if your application config maps the new variable.

Pricing

Type	Price (per 1M)
Input tokens	$0.05
Output tokens	$0.25

Capabilities

No model capability flags are currently sourced.

About Stable LM 3B

Stable LM 3B is a 3-billion parameter language model from Stability AI, designed for efficient operation on devices like smartphones and laptops. This compact model reduces operating costs and environmental impact while achieving performance comparable to larger open-source models. Its decoder-only transformer architecture features Rotary Position Embeddings and LayerNorm with learned bias terms. Trained on 1 trillion tokens from diverse sources, the model delivers strong conversational abilities but may require fine-tuning and safety testing for specific applications. An instruction-fine-tuned version was planned, emphasizing its potential for safe deployment.