Llama 3.2 3B is a 3-billion-parameter multilingual large language model, optimized for advanced natural language processing tasks like dialogue generation, reasoning, and summarization. Designed with the latest transformer architecture, it...
- Modalities
- text → text
- In / out per 1M
- Free / Free
- Context
- 131K tokens
- Added
- Sep 25, 2024
- Tokenizer
- Llama3
- Knowledge cutoff
- Dec 31, 2023
Pricing
Per 1M tokens. The provider price and our flat 3% fee are separate columns — what you pay is their sum.
| Per 1M tokens | Provider | + 3% fee | You pay |
|---|---|---|---|
| Input | Free | — | Free |
| Output | Free | — | Free |
Providers
No provider breakdown is published for Llama 3.2 3B Instruct (free) in the current catalog snapshot (Jul 28, 2026). Browse all providers
Supported parameters
- frequency_penalty
- max_tokens
- presence_penalty
- stop
- temperature
- top_k
- top_p
Call it
OpenAI-compatible: point your SDK at api.openkey.ai/v1 and use
model meta-llama/llama-3.2-3b-instruct:free.
curl https://api.openkey.ai/v1/chat/completions \
-H "Authorization: Bearer ***" \
-H "Content-Type: application/json" \
-d '{
"model": "meta-llama/llama-3.2-3b-instruct:free",
"messages": [{"role": "user", "content": "Hello"}]
}'import os
from openai import OpenAI
client = OpenAI(
base_url="https://api.openkey.ai/v1",
api_key=os.environ["OPENKEY_API_KEY"],
)
completion = client.chat.completions.create(
model="meta-llama/llama-3.2-3b-instruct:free",
messages=[{"role": "user", "content": "Hello"}],
)
print(completion.choices[0].message.content)import OpenAI from "openai";
const client = new OpenAI({
baseURL: "https://api.openkey.ai/v1",
apiKey: process.env.OPENKEY_API_KEY,
});
const completion = await client.chat.completions.create({
model: "meta-llama/llama-3.2-3b-instruct:free",
messages: [{ role: "user", content: "Hello" }],
});
console.log(completion.choices[0].message.content);About Llama 3.2 3B Instruct (free)
Llama 3.2 3B Instruct (free) is Meta's 3-billion-parameter text-to-text model, built on the Llama3 tokenizer with a 131,072-token context window and a knowledge cutoff of December 31, 2023. At this size, it's built for lightweight dialogue, summarization, and reasoning tasks rather than heavy multi-step work — no benchmark scores or reasoning-mode support are published for it. It takes text in, returns text out, and supports standard sampling controls like temperature, top_p, and top_k, plus frequency and presence penalties.
It's one of 25 free models on OpenKey, priced at $0 for both input and output versus its paid sibling Llama 3.2 3B Instruct at $0.0509/$0.335 per 1M tokens. Its 131,072-token context window beats 23% of the catalog's models — modest for context, but usable for documents in the tens-of-pages range. There's no vision input, no tool-calling parameter listed, and no benchmark data, so it's a plain-text, low-cost option rather than a multimodal or agentic pick.
Questions
- How much does Llama 3.2 3B Instruct (free) cost via API?
- It's free — $0 per 1M input tokens and $0 per 1M output tokens, both on the provider side and through OpenKey. Even with OpenKey's flat 3% fee applied to provider pricing ($0 x 1.03 = $0), there's no cost. If you need more headroom later, the paid Llama 3.2 3B Instruct sibling runs $0.0509 input / $0.335 output per 1M tokens.
- What is Llama 3.2 3B Instruct (free)'s context window?
- The context window is 131,072 tokens. That's enough to hold a long report, a small codebase, or a few hundred pages of text in a single request. It ranks larger than 23% of models in the OpenKey catalog, so it's on the smaller end for context despite the six-figure token count.
- Is Llama 3.2 3B Instruct (free) free to use?
- Yes, it's fully free — both prompt and completion tokens are priced at $0 per 1M. It's one of 25 free models available on OpenKey out of 329 total models. There's no fee markup either, since 3% of $0 is still $0.
- Does Llama 3.2 3B Instruct (free) support tool calling or vision?
- No — it's text-in, text-out only, with no vision or image input modality listed. Its supported parameters cover frequency_penalty, max_tokens, presence_penalty, stop, temperature, top_k, and top_p, but tool calling isn't among them. For vision, Meta's own Llama 3.2 11B Vision Instruct (priced at $0.345 per 1M tokens on both input and output) is the sibling to use instead.
- What is Llama 3.2 3B Instruct (free)'s knowledge cutoff?
- Its training data goes up to December 31, 2023. That means it won't know about events, releases, or developments from 2024 onward. For anything more recent, you'd need to supply the context yourself or pick a model with a later cutoff.
More from Meta AI
All Meta AI models →
Meta AILlama Guard 4 12B
Llama Guard 4 is a Llama 4 Scout-derived multimodal pretrained model, fine-tuned for content safety classification. Similar to previous versions, it can be used to classify content in both LLM...
ctx 164K in $0.185 out $0.185
Meta AILlama 4 Maverick
Llama 4 Maverick 17B Instruct (128E) is a high-capacity multimodal language model from Meta, built on a mixture-of-experts (MoE) architecture with 128 experts and 17 billion active parameters per forward...
ctx 1.0M in $0.154 out $0.618
Meta AILlama 4 Scout
Llama 4 Scout 17B Instruct (16E) is a mixture-of-experts (MoE) language model developed by Meta, activating 17 billion parameters out of a total of 109B. It supports native multimodal input...
ctx 10M in $0.103 out $0.309
Meta AILlama 3.3 70B Instruct
The Meta Llama 3.3 multilingual large language model (LLM) is a pretrained and instruction tuned generative model in 70B (text in/text out). The Llama 3.3 instruction tuned text only model...
ctx 131K in $0.103 out $0.330