Llama 3.3 70B Instruct pricing.
Prices checked
Llama 3.3 70B Instruct costs $0.10 in and $0.32 out per million tokens from OpenRouter; the cheapest price is $0.10 at NanoGPT, checked .
OpenRouter's own price
OpenRouter
$0.10 in · $0.32 out
82% below the typical price of the 392 text models we track.
Cheapest price
NanoGPT
$0.05 in · $0.23 out
39% less than OpenRouter, and the provider says why.
Every provider selling Llama 3.3 70B Instruct
- 1
$0.05 in · $0.23 out
$0.10/M tokens - 2
$0.10 in · $0.30 out
$0.15/M tokens - 3
$0.10 in · $0.32 out
$0.16/M tokens - 4
$0.10 in · $0.32 out
$0.16/M tokens - 5
$0.10 in · $0.32 out
$0.16/M tokens - 6
$0.10 in · $0.32 out
$0.16/M tokens - 7
$0.10 in · $0.32 out
$0.16/M tokens - 8
$0.10 in · $0.32 out
$0.16/M tokens - 9
$0.11 in · $0.34 out
$0.16/M tokens - 10
$0.12 in · $0.30 out
$0.16/M tokens - 11
$0.12 in · $0.30 out
$0.16/M tokens - 12
$0.13 in · $0.39 out · 128k context here
$0.20/M tokens - 13
$0.13 in · $0.39 out
$0.20/M tokens - 14
$0.13 in · $0.40 out
$0.20/M tokens - 15
$0.13 in · $0.40 out
$0.20/M tokens - 16
$0.20 in · $0.20 out
$0.20/M tokens - 17
$0.14 in · $0.40 out
$0.20/M tokens - 18
$0.14 in · $0.40 out · 12k context here
$0.20/M tokens - 19
$0.14 in · $0.40 out · 12k context here
$0.20/M tokens - 20
$0.14 in · $0.42 out
$0.21/M tokens - 21
$0.23 in · $0.40 out
$0.27/M tokens - 22
$0.23 in · $0.40 out
$0.27/M tokens - 23
AkashML
Costs more$0.20 in · $0.52 out
$0.28/M tokens - 24
$0.22 in · $0.50 out
$0.29/M tokens - 25
$0.22 in · $0.50 out
$0.29/M tokens
- 26
$0.22 in · $0.50 out
$0.29/M tokens - 27
$0.25 in · $0.75 out · 128k context here
$0.38/M tokens - 28
$0.58 in · $0.77 out · 128k context here
$0.63/M tokens - 29
$0.59 in · $0.79 out
$0.64/M tokens - 30
$0.59 in · $0.79 out
$0.64/M tokens - 31
$0.70 in · $0.70 out · 128k context here
$0.70/M tokens - 32
$0.71 in · $0.71 out · 128k context here
$0.71/M tokens - 33
CoreWeave
Costs more$0.71 in · $0.71 out · 128k context here
$0.71/M tokens - 34
$0.71 in · $0.71 out · 128k context here
$0.71/M tokens - 35
CoreWeave
Costs more$0.71 in · $0.71 out · 128k context here
$0.71/M tokens - 36
$0.71 in · $0.71 out · 128k context here
$0.71/M tokens - 37
$0.72 in · $0.72 out · 128k context here
$0.72/M tokens - 38
$0.72 in · $0.72 out · 128k context here
$0.72/M tokens - 39
$0.73 in · $0.73 out
$0.73/M tokens - 40
$0.73 in · $0.73 out · 128k context here
$0.73/M tokens - 41
$0.74 in · $0.74 out
$0.74/M tokens - 42
$0.60 in · $1.20 out · 128k context here
$0.75/M tokens - 43
watsonx.ai
Costs more$0.75 in · $0.75 out
$0.75/M tokens - 44
$0.29 in · $2.25 out · 24k context here
$0.78/M tokens - 45
$0.29 in · $2.25 out · 24k context here
$0.78/M tokens - 46
$0.29 in · $2.25 out · 24k context here
$0.78/M tokens - 47
$0.31 in · $2.38 out · 24k context here
$0.83/M tokens - 48
$0.85 in · $1.20 out · 128k context here
$0.94/M tokens - 49
$1.01 in · $1.01 out · 100k context here
$1.01/M tokens - 50
$1.04 in · $1.04 out
$1.04/M tokens - 51
Regolo AI
Costs more$0.60 in · $2.70 out · 128k context here
$1.13/M tokens - 52
Models Think Evroc
Costs more$1.15 in · $1.15 out · 128k context here
$1.15/M tokens - 53Costs more
$0.76 in · $2.38 out · 128k context here
$1.17/M tokens - 54
$0.70 in · $2.80 out · 128k context here
$1.22/M tokens - 55
GreenPT
Costs more$1.25 in · $1.25 out · 100k context here
$1.25/M tokens - 56Costs more
$1.04 in · $2.25 out
$1.34/M tokens - 57
$2 in · $2 out
$2/M tokens - 58
CloudFerro Sherlock
Costs more$2.92 in · $2.92 out · 70k context here
$2.92/M tokens
Not ranked
Below the maker's own price with nothing published to account for it, from a catalog we could not read reliably, or a free tier, which is a limit rather than a price. Why
$0.45 in · $0.90 out
$0.56/M tokens$0.65 in · $0.75 out · 33k context here
$0.68/M tokens
What it costs for your work
A month at OpenRouter's own price
A month at NanoGPT, the cheapest
What you save
A page is about 600 tokens in and 60 out, a support reply 200 in and 300 out, an email 150 in and 250 out. Nothing else is counted: no cache discount, no batch rate, no free tier.
Run it yourself
On a server by the month
Meta publishes the weights, so you can run Llama 3.3 70B Instruct on a server you rent by the month. It needs about 46.5 GB of RAM on a CPU, and the cheapest plan we track with that much and a disk included is Leaseweb VPS 5 at $34.18 a month, streaming about 0.6–1.1 tok/s. Every plan that fits
On a GPU by the hour
The RTX A6000 holds Llama 3.3 70B Instruct in 48 GB of memory against the 46.5 GB the weights, the context and the runtime need. The cheapest one we track is RunPod's RTX A6000 · Community Cloud at $0.33/GPU-hr, which is $240.90 for the 730 hours in a month, or $0.41/GPU-hr interruptible, checked 31 Oct. Every price for the RTX A6000
How we count
Prices are per million tokens. The big number on each row is blended 3 in : 1 out, the way most work uses a model. A price below OpenRouter's own is listed as cheapest only when the provider says why. Meta trained Llama 3.3 70B Instruct and sells it on no catalog we read, so the price every row is measured against is OpenRouter's.
- Resells
- The provider sells another company's endpoint, at a margin of its own or at the maker's price.
- Same as maker
- Sold by someone else at the price the maker charges.
- Costs more
- Sold by someone else for more than the maker charges.
- Not ranked
- Below the maker's price on a closed model with nothing to account for it, or from a catalog we could not read reliably. Listed, and never a pick.
Similar models
Models from other labs that do what Llama 3.3 70B Instruct does, cheapest first. Each percentage is against Llama 3.3 70B Instruct's cheapest price, $0.10 per M tokens.
gpt-oss-120b
131k context · 73 providers
Kilo Gateway$0.06/M tokens37% lessGemma 3 27B
131k context · 23 providers
Hugging Face Inference Providers$0.10/M tokens5% moreMistral Small 3.2
128k context · 25 providers
Helicone$0.11/M tokens12% moreQwen3 32B
131k context · 29 providers
OpenRouter$0.13/M tokens37% more
Frequently asked
- Who sells Llama 3.3 70B Instruct cheapest?
- NanoGPT, at $0.05 in and $0.23 out per million tokens, checked 31 Oct 2026. The provider sells another company's endpoint, at a margin of its own or at the maker's price.
- Is Llama 3.3 70B Instruct open weights?
- Yes. Meta publishes the weights, so anyone may host Llama 3.3 70B Instruct, which is why 50 providers sell it and why their prices differ so much.
- How much context does Llama 3.3 70B Instruct hold?
- 131,072 tokens in one call, which is about 131k, and it can call your own functions.
- How does Llama 3.3 70B Instruct compare with the rest of the market?
- At $0.16 per million tokens, OpenRouter's own price is 82% below the typical $0.88 across the 392 text models we track.