Model tracker / Meta
Llama 3.3 70B
Llama 3.3 70B is a fast open-weight model from Meta, released 6 December 2024. At DeepInfra it costs $0.10 per million input tokens and $0.32 per million output tokens, with a 128K-token context window. 6 providers serve it, and DeepInfra is the cheapest.
- Made by
- Meta Platforms
- Released
- 2024-12-06
- Status
- Generally available
- Tier
- Fast
- Weights
- Open
- Context window
- 128,000 tokens
- Max output
- 16,384 tokens
- Input
- text
- API id
Llama-3.3-70B-Instruct-Turbo
Llama 3.3 70B price at 6 providers
US dollars per million tokens. Blended is 3 input tokens for every output token, one number to compare on. The cheapest blended price is highlighted.
| Provider | Input | Output | Cached input | Blended | Batch | Checked |
|---|---|---|---|---|---|---|
| DeepInfra | $0.10 | $0.32 | – | $0.155 | – | 2026-10-15 |
| Novita AI · meta-llama/llama-3.3-70b-instruct | $0.135 | $0.40 | – | $0.201 | – | 2026-10-15 |
| Microsoft · Llama-3.3-70B-Instruct | $0.71 | $0.71 | – | $0.71 | – | 2026-10-15 |
| SambaNova · Meta-Llama-3.3-70B-Instruct | $0.60 | $1.20 | – | $0.75 | – | 2026-10-15 |
| Cloudflare · @cf/meta/llama-3.3-70b-instruct-fp8-fast | $0.293 | $2.25 | – | $0.783 | – | 2026-10-15 |
| Together AI · meta-llama/Llama-3.3-70B-Instruct-Turbo | $1.04 | $1.04 | – | $1.04 | – | 2026-10-15 |
- Novita AI: Novita serves it with a 12K context window.
- Microsoft: Azure AI Foundry, Global deployment. Data Zone and regional deployments cost more, from $0.781 to $0.937 per million tokens each way depending on region.
- Cloudflare: FP8-quantized fast variant with a 24K context window. Billed in Neurons at $0.011 per 1,000; 10,000 Neurons a day free.
- Together AI: Scheduled for removal from Together serverless on 2026-10-22; Together recommends Muse Glimmer 30B instead.
Models like Llama 3.3 70B
Other fast models closest in price, at each maker's own price or the cheapest provider.
| Model | Input | Output | Context | Weights |
|---|---|---|---|---|
| Llama 3.3 70B | $0.10 | $0.32 | 128K | Open |
| Ministral 3 8B Mistral AI | $0.15 | $0.15 | 256K | Open |
| NVIDIA Nemotron 3 Super NVIDIA | $0.085 | $0.40 | 256K | Open |
| Claude Haiku 5.5 Anthropic | $0.10 | $0.50 | 1M | Closed |
| Ministral 3 14B Mistral AI | $0.20 | $0.20 | 256K | Open |
| GPT-6 Luna OpenAI | $0.10 | $0.50 | 1.1M | Closed |
Compare them all in the model tracker →
Llama 3.3 70B FAQ
How much does Llama 3.3 70B cost?
Llama 3.3 70B costs $0.10 per million input tokens and $0.32 per million output tokens at DeepInfra. Prices are list prices in US dollars, last checked 15 October 2026.
What is Llama 3.3 70B's context window?
Llama 3.3 70B takes up to 128,000 tokens in one request and writes up to 16,384 tokens in one response.
Is Llama 3.3 70B open weight?
Yes. Meta publishes the weights, so you can run Llama 3.3 70B yourself or at any host that serves it.
Where can I use Llama 3.3 70B?
Llama 3.3 70B is available from DeepInfra, Novita AI, Microsoft, SambaNova, Cloudflare and Together AI. At DeepInfra, call it as Llama-3.3-70B-Instruct-Turbo.