Llama 3.3 70B
llama-3.3-70bPricing
List prices at Inference APIs, in US dollars per million tokens.
| Tier | Input / 1M | Output / 1M | |
|---|---|---|---|
| Standard | $0.20 | $0.50 |
What it costs a month
At Inference APIs list prices, 30 days.
- Chat assistant5,000 requests a day, 2K in / 0.5K out$97.50
- RAG over documents2,000 requests a day, 8K in / 0.6K out$114
- Coding agent1,000 requests a day, 40K in / 2K out$270
- Overnight batch20,000 requests a day, 4K in / 0.4K out$600
Where to run it
4 hosts serve Llama 3.3 70B. The comparison is on blended price, 3 parts input to 1 part output.
| Host | In / out per 1M | Model id | |
|---|---|---|---|
Kyma APICloud llama-3.3-70b | $0.135 / $0.432 −24% vs headline | llama-3.3-70b | |
Inference APIsGateway llama-3.3-70b | $0.20 / $0.50 | llama-3.3-70b | |
Vercel AI GatewayGateway llama-3.3-70b | $0.72 / $0.72 +162% vs headline | llama-3.3-70b | |
CerebrasInference llama-3.3-70b | $0.85 / $1.20 +241% vs headline | llama-3.3-70b |
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 LicenseRuntimes: Ollama, TGI, llama.cpp, vLLM
Cheaper alternatives
Current models that cost less and keep most of what this one can do: tools, vision, reasoning, structured output, caching and enough context.
Same model, cheaper host
- Kyma API24% lessllama-3.3-70b$0.135 / $0.432
Smaller Meta models
- 55% less
- 27% less
Other frontier labs
- 27% less
- 27% less
- 5% less
Open-weight picks
- 5% less
- 8% less
- 35% less
- 42% less
Model line
Newer and older versions, at each lab's own price.
- UndatedLlama 3.3 70BLatest$0.20 / $0.50
- UndatedLlama3.3 70B$0.72 / $0.72
- UndatedLlama 3.1 70B$0.60 / $0.60
- UndatedLlama3.1 70B$0.72 / $0.72
- UndatedLlama 3 70B$0.59 / $0.79
- UndatedLlama3 70BNot listed
- UndatedLlama 2 70B$0.10 / $0.10
- UndatedLlama2 70BNot listed
Call it
Using Inference APIs's API.
from litellm import completion
response = completion(
model="inferenceapis/llama-3.3-70b",
messages=[{"role": "user", "content": "Hello"}],
)
print(response.choices[0].message.content)Quick answers
How much does Llama 3.3 70B cost?
On Inference APIs, Llama 3.3 70B costs $0.20 per million input tokens and $0.50 per million output tokens. The cheapest host is Kyma API, at $0.135 per million input tokens and $0.432 per million output tokens.
Which APIs offer Llama 3.3 70B?
Llama 3.3 70B runs on 4 APIs: Kyma API, Inference APIs, Vercel AI Gateway and Cerebras.
What is Llama 3.3 70B's context window?
131K tokens, with up to 131K tokens in one reply.
Is Llama 3.3 70B being retired?
No retirement date is scheduled for it on any API we track.
Is there a newer version of Llama 3.3 70B?
No. Llama 3.3 70B is the newest model in its line.
What are cheaper alternatives to Llama 3.3 70B?
Models that cost less and keep most of what it can do: Muse Spark 1.3 Contributor on Meta Model API (55% less), Claude Haiku 5.5 on Anthropic (27% less), GPT-6 Luna on OpenAI (27% less) and Mistral Small 2603 on Mistral AI (5% less). Llama 3.3 70B itself costs 24% less on Kyma API. Savings are on a mix of 3 input tokens to 1 output token.
What is Llama 3.3 70B's API model id?
Use llama-3.3-70b with Inference APIs.
Sources
Raw catalog entry for llama-3.3-70b on Inference APIs
{
"input_cost_per_token": 2e-7,
"litellm_provider": "inferenceapis",
"max_input_tokens": 131072,
"max_output_tokens": 131072,
"mode": "chat",
"output_cost_per_token": 5e-7,
"source": "https://inferenceapis.com/pricing",
"supported_endpoints": [
"/v1/chat/completions",
"/v1/responses",
"/v1/messages"
],
"supports_function_calling": true,
"supports_native_streaming": true,
"supports_reasoning": false,
"supports_tool_choice": true,
"supports_vision": false
}