impulse

Llama 3.3 70B Instruct

Open weights

The Meta Llama 3.3 multilingual large language model (LLM) is a pretrained and instruction tuned generative model in 70B (text in/text out).

API id Llama-3.3-70B-InstructReleased 6 Dec 2024 (Dec 2024)
Azure AI Foundry retires it on 31 Jan 2027 (in 74 days). 15 hosts keep serving it.
Input / 1M
-
not listed
Output / 1M
-
not listed
Context window
128K
tokens
Max output
4K
tokens per reply
Hosts
16
17 listings
Released
Dec 2024
6 Dec 2024

Where to run it

16 hosts serve Llama 3.3 70B Instruct. The comparison is on blended price, 3 parts input to 1 part output.

HostIn / out per 1M
Llama APILab API
Llama-3.3-70B-Instruct
-
HyperbolicInference
Llama-3.3-70B-Instruct
$0.12 / $0.30
Llama-3.3-70B-Instruct
$0.13 / $0.40
CrusoeInference
Llama-3.3-70B-Instruct
$0.20 / $0.20
NscaleInference
Llama-3.3-70B-Instruct
$0.20 / $0.20
Novita AIInference
llama-3.3-70b-instruct
$0.135 / $0.40
DeepInfraInference
Llama-3.3-70B-Instruct
$0.23 / $0.40
Llama-3.3-70B-Instruct
$0.20 / $0.50
GMI CloudInference
Llama-3.3-70B-Instruct
$0.25 / $0.75
Llama-3.3-70B-Instruct
$0.71 / $0.71
Retires 31 Jan 2027
W&B InferenceInference
Llama-3.3-70B-Instruct
$0.71 / $0.71
llama-3.3-70b-instruct-maas
$0.72 / $0.72
meta.llama-3.3-70b-instruct
$0.72 / $0.72
llama-3-3-70b-instruct
$0.753 / $0.753
llama-3.3-70b-instruct
$0.90 / $0.90
Fireworks AIInference
llama-v3p3-70b-instruct
$0.90 / $0.90
Show 1 more listings: regions, snapshots and variants
Inference APIs
Alias
llama-3.3-70b-instruct
$0.20 / $0.50

Capabilities

Core

  • Streaming
  • Function calling
  • JSON mode
  • Structured outputs

Multimodal

  • Vision
  • Image input
  • Audio
  • Image generation
  • Video

Advanced

  • Reasoning
  • Tool calling
  • MCP
  • Prompt caching
  • Batch API
  • Embeddings
Takes in
Text
Produces
Text
Open weights
License: Llama Community License
Runs in about 169.3 GB of GPU memory
Runtimes: Ollama, TGI, llama.cpp, vLLM

Similar models

This model has no listed price, so these are current models with similar capabilities.

Smaller Meta models

Other frontier labs

  • GPT-5.5 Pro
    OpenAI · on OpenAI
    $30.00 / $180.00
    Has every key capability
  • Claude Fable 5.2
    Anthropic · on Anthropic
    $8.00 / $40.00
    Has every key capability
  • Gemini 4 Carbon
    Google · on Google AI Studio
    $2.00 / $10.00
    Has every key capability
  • Grok 4.7
    xAI · on xAI
    $2.00 / $6.00
    Has every key capability

Open-weight picks

  • Command A Plus
    Cohere · on AIHubMix
    $2.50 / $10.00
    Has every key capability
  • Kimi K3.1 Pro
    Moonshot AI · on Databricks
    $2.00 / $10.00
    Has every key capability
  • Qwen3.8.2 4T A95B
    Alibaba Qwen · on DeepInfra
    $2.00 / $6.00
    Has every key capability
  • Muse Spark Open
    Meta · on Baseten
    $1.25 / $4.25
    Has every key capability

Benchmarks

Scores published on the model card.

  • BBH56.6%
  • GPQA10.5%
  • IFEval90.0%
  • MATH Lvl 548.3%
  • MMLU-Pro48.1%
  • MUSR15.6%
  • Open LLM Average44.8%

Model line

Newer and older versions, at each lab's own price.

  1. Dec 2024Llama 3.3 70B InstructLatestNot listed
  2. Dec 2024Llama3.3 70B Instruct$0.12 / $0.30
  3. Jul 2024Llama 3.1 70B Instruct$0.72 / $0.72
  4. Jul 2024Llama3.1 70B Instruct$0.12 / $0.30
  5. UndatedLlama 3 70B Instruct$0.51 / $0.74
  6. UndatedLlama3 70B Instruct$2.65 / $3.50

Call it

Using Llama API's API.

Python
from litellm import completion

response = completion(
    model="meta_llama/Llama-3.3-70B-Instruct",
    messages=[{"role": "user", "content": "Hello"}],
)
print(response.choices[0].message.content)

Quick answers

How much does Llama 3.3 70B Instruct cost?

Its lowest listed price is at Hyperbolic: $0.12 per million input tokens and $0.30 per million output tokens.

Which APIs offer Llama 3.3 70B Instruct?

Llama 3.3 70B Instruct runs on 16 APIs: Llama API, Hyperbolic, Nebius AI Studio, Crusoe, Nscale, Novita AI, DeepInfra, Inference APIs, GMI Cloud and 7 more.

What is Llama 3.3 70B Instruct's context window?

128K tokens, with up to 4K tokens in one reply.

Is Llama 3.3 70B Instruct being retired?

Azure AI Foundry retires it on 31 Jan 2027. 15 hosts keep serving it.

Is there a newer version of Llama 3.3 70B Instruct?

No. Llama 3.3 70B Instruct is the newest model in its line.

When was Llama 3.3 70B Instruct released?

Meta released it on 6 Dec 2024.

What is Llama 3.3 70B Instruct's API model id?

Use Llama-3.3-70B-Instruct with Llama API. Other hosts use their own ids: llama-3.3-70b-instruct on Novita AI, llama-3.3-70b-instruct-maas on Vertex AI, meta.llama-3.3-70b-instruct on Oracle OCI, llama-3-3-70b-instruct on IBM watsonx.ai.

Sources

Raw catalog entry for Llama-3.3-70B-Instruct on Llama API
{
  "litellm_provider": "meta_llama",
  "max_input_tokens": 128000,
  "max_output_tokens": 4028,
  "mode": "chat",
  "source": "https://ai.developer.meta.com/docs/models",
  "supported_modalities": [
    "text"
  ],
  "supported_output_modalities": [
    "text"
  ],
  "supports_function_calling": true,
  "supports_tool_choice": true
}