All models
Meta
Llama 3.3 70B Instruct fastruns locally
meta/llama-3-3-70b
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...
Input / 1M
$0.20
USD per million tokens
Output / 1M
$0.52
USD per million tokens
Est. cost / token
$0.00000028
$0.28 per 1M, 3:1 mix
Context window
131K
max output 128K
Major features
Open weightsCachingJSON schemaTools
Task suitability (heuristic 1–5, not a benchmark)
- coding3
- reasoning2
- math2
- summarization4
- extraction4
- classification4
- chat3
- creative3
- translation4
- vision1
- long context3
- agentic3
Run it locally
Will it run on my machine? →ollama run llama3.3:70b| Quantization | Quality | Memory needed (8K ctx) |
|---|---|---|
| Q2_K | poor | 34.3 GB |
| Q3_K_M | fair | 39.5 GB |
| Q4_K_Mrecommended | good | 48.6 GB |
| Q5_K_M | very good | 56 GB |
| Q6_K | excellent | 64.1 GB |
| Q8_0 | near-lossless | 82 GB |
| F16 | lossless | 152 GB |
Estimates.
Price history
| Recorded | Input | Output | Cached | Source |
|---|---|---|---|---|
| 2026-11-05 03:20 | $0.20 | $0.52 | $0.10 | openrouter |
A row is added whenever a sync detects a price change. Vendor pricing source
Pricing (USD per 1M tokens)
- Input
- $0.20
- Output
- $0.52
- Cached input
- $0.10
- Blended (3:1)
- $0.28
Details
- Company
- Meta
- Version
- 3.3
- Family
- Llama
- Released
- 2024-12-06
- Parameters
- 70.6B
- Max output
- 128K
- Open weights
- Yes
- Inputs
- text
- Last synced
- 46 min ago