Llama-3.3-70B-Instruct

meta/llama-3.3-70b-instruct

Maker: Meta, United States
Released: 2024-12-06
Knowledge cutoff: 2023-12
Family: llama
Weights: open (llama3.3)
Tasks: text.chat
Cheapest offering: DeepInfra, about $0.000196 for a typical text.chat call

Try it in the sandbox Use it from the API

Offerings

Provider	Model id at the provider	Status	Price per 1M tokens	Per-unit prices	Limits	Capabilities (level)
Azure OpenAI	`llama-3.3-70b-instruct`	ga	Input: $0.71 Output: $0.71 Since 2026-11-04 · models.dev, MIT	none	Context: 128,000 tokens Max output: 32,768 tokens	`input.text` claimed `output.text` claimed `tools` claimed
Amazon Bedrock	`meta.llama3-3-70b-instruct-v1:0`	deprecateddeprecated 2027-03-30	Input: $0.72 Output: $0.72 Since 2026-11-03 · models.dev, MIT	none	Context: 128,000 tokens Max output: 4,096 tokens	`input.text` claimed `output.text` claimed `tools` claimed
Amazon Bedrock	`us.meta.llama3-3-70b-instruct-v1:0`	deprecateddeprecated 2027-03-30	Input: $0.72 Output: $0.72 Since 2026-11-03 · models.dev, MIT	none	Context: 128,000 tokens Max output: 4,096 tokens	`input.text` claimed `output.text` claimed `tools` claimed
DeepInfra	`meta-llama/Llama-3.3-70B-Instruct-Turbo`	ga	Input: $0.1 Output: $0.32 Since 2026-11-03 · models.dev, MIT	none	Context: 131,072 tokens Max output: 16,384 tokens	`input.text` claimed `output.text` claimed `tools` claimed `structured_output` claimed `tool_choice` claimed
Groq	`llama-3.3-70b-versatile`	deprecated	Input: $0.59 Output: $0.79 Since 2026-11-03 · models.dev, MIT	none	Context: 131,072 tokens Max output: 32,768 tokens	`input.text` claimed `output.text` claimed `tools` claimed
OpenRouter	`meta-llama/llama-3.3-70b-instruct`	ga	Input: $0.2 Output: $0.52 Cache read: $0.1 Unverified: sources disagree by more than 5%Since 2026-11-04 · models.dev, MIT	none	Context: 131,072 tokens Max output: 128,000 tokens	`input.text` claimed `output.text` claimed `tools` claimed `structured_output` claimed
Together AI	`meta-llama/Llama-3.3-70B-Instruct-Turbo`	deprecateddeprecated 2026-10-22	Input: $1.04 Output: $1.04 Unverified: sources disagree by more than 5%Since 2026-11-03 · models.dev, MIT	none	Context: 131,072 tokens Max output: 131,072 tokens	`input.text` claimed `output.text` claimed `tools` claimed `parallel_tools` claimed `tool_choice` claimed `structured_output` claimed

Run it locally

Free on your own device. Memory is the weights, the KV cache at the context shown and about 10% overhead; a longer context needs more. GB are 10⁹ bytes. See local models.

Runtime	Model id in the runtime	Quantization	Needs	Context
llama.cpp	`lmstudio-community/Llama-3.3-70B-Instruct-GGUF:Q4_K_M`	Q4_K_M	Memory: 49.73 GB at 8,192 tokens GPU memory: 49.73 GB Disk: 42.53 GB Parameters: 70.6B License: llama3.3	Context: 131,072 tokens
llama.cpp	`lmstudio-community/Llama-3.3-70B-Instruct-GGUF:Q6_K`	Q6_K	Memory: 66.63 GB at 8,192 tokens GPU memory: 66.63 GB Disk: 57.89 GB Parameters: 70.6B License: llama3.3	Context: 131,072 tokens
llama.cpp	`lmstudio-community/Llama-3.3-70B-Instruct-GGUF:Q8_0`	Q8_0	Memory: 85.43 GB at 8,192 tokens GPU memory: 85.43 GB Disk: 74.98 GB Parameters: 70.6B License: llama3.3	Context: 131,072 tokens
LM Studio	`lmstudio-community/Llama-3.3-70B-Instruct-GGUF/Llama-3.3-70B-Instruct-Q4_K_M.gguf`	Q4_K_M	Memory: 49.73 GB at 8,192 tokens GPU memory: 49.73 GB Disk: 42.53 GB Parameters: 70.6B License: llama3.3	Context: 131,072 tokens
LM Studio	`lmstudio-community/Llama-3.3-70B-Instruct-GGUF/Llama-3.3-70B-Instruct-Q6_K-00001-of-00002.gguf`	Q6_K	Memory: 66.63 GB at 8,192 tokens GPU memory: 66.63 GB Disk: 57.89 GB Parameters: 70.6B License: llama3.3	Context: 131,072 tokens
LM Studio	`lmstudio-community/Llama-3.3-70B-Instruct-GGUF/Llama-3.3-70B-Instruct-Q8_0-00001-of-00002.gguf`	Q8_0	Memory: 85.43 GB at 8,192 tokens GPU memory: 85.43 GB Disk: 74.98 GB Parameters: 70.6B License: llama3.3	Context: 131,072 tokens
MLX	`mlx-community/Llama-3.3-70B-Instruct-4bit`	4bit	Memory: 46.62 GB at 8,192 tokens GPU memory: 46.62 GB Disk: 39.69 GB Parameters: 70.6B License: llama3.3	Context: 131,072 tokens
Ollama	`llama3.3:70b`	Q4_K_M	Memory: 50.81 GB at 8,192 tokens GPU memory: 50.81 GB Disk: 43.5 GB Parameters: 70.6B License: llama3.3	Context: 131,072 tokens

Quality scores

Basis	Score (0–1)	Raw	Measured	Source
`curated`	0.400 (estimated)	0.4000	2026-11-02	Curated tier (estimated), Nanner Tech

Aliases

Other names the API accepts for this model.

llama-3.3-70b-versatile
meta-llama/llama-3.3-70b-instruct
meta-llama/llama-3.3-70b-instruct-turbo
meta.llama3-3-70b-instruct-v1:0
us.meta.llama3-3-70b-instruct-v1:0

Use it from the API

Look this model up, or ask for the best model for its task on the providers that offer it. See the quickstart and the API reference.

Look up this model

curl 'https://rhumbdata.com/v1/models/meta/llama-3.3-70b-instruct' \
  -H "Authorization: Bearer $RHUMB_KEY"

Resolve text.chat on its providers

curl 'https://rhumbdata.com/v1/resolve?task=text.chat&providers=azure-openai,bedrock,deepinfra,groq,openrouter,together' \
  -H "Authorization: Bearer $RHUMB_KEY"

Built from catalog version 2026-11-04.3 of 2026-11-04. Prices are list prices per 1M tokens unless marked otherwise; every score shows its source's notice (see Attributions).