Llama 3.2 3B API pricing
9 providers serve Llama 3.2 3B. Llama 3.2 3B is an open-weight model (Llama 3.2 Community) with 3.2B total parameters, up to 131K context. Prices below are per 1M tokens, cheapest input first (self-hosting math further down).
Refreshed 2026-08-23 · list prices from provider APIs| Provider | Input $/1M | Output $/1M | Cache read $/1M | Context | |
|---|---|---|---|---|---|
| DeepInfra | $0.020 | $0.020output floor | - | 131K | Use |
| Novita AI | $0.030 | $0.050 | - | 33K | Use |
| Llamagate | $0.040 | $0.080 | - | 131K | |
| OpenRouter marketplace quote | $0.050 | $0.330 | - | 131K | Use |
| Cloudflare | $0.051 | $0.335 | - | 80K | Use |
| SambaNova | $0.080 | $0.160 | - | 4K | Use |
| Hyperbolic | $0.120 | $0.300 | - | 33K | Use |
| Vercel AI Gateway | $0.150 | $0.150 | - | 128K | Use |
| IBM watsonx | $0.150 | $0.150 | - | 128K | Use |
The cheapest way to run Llama 3.2 3B
On an illustrative 70M-input / 30M-output monthly workload, today's lowest listed cost is roughly $2.00/month, at DeepInfra. Input-heavy, output-heavy and cache-heavy workloads can produce different winners. Use the live mix calculator below rather than combining floors from two different providers. On output-token cost alone, a busy self-hosted deployment can work out cheaper (~$0.0058/1M output on V100). See the math below. Run your own numbers in the breakeven calculator, or read the breakeven math.
Your actual monthly cost
Cache read uses the listed cache rate where available; otherwise it falls back to normal input price. Batch, write-cache, volume and negotiated discounts are excluded.
Or run it yourself: Llama 3.2 3B self-hosting economics
Llama 3.2 3B is open-weight (Llama 3.2 Community), so the API price above competes with the GPU-hour market. 3.2B total parameters needs roughly 5 GB VRAM at FP8 or 3 GB at INT4, KV-cache headroom included.
| GPU that fits (FP8, single card) | VRAM | From $/hr | Cheapest at | |
|---|---|---|---|---|
| V100 | 32 GB | $0.022marketplace | Vast.ai | Rent |
| RTX 3090 | 24 GB | $0.068marketplace | Vast.ai | Rent |
| RTX 5080 | 16 GB | $0.102marketplace | Vast.ai | Rent |
| RTX A2000 | 6 GB | $0.120community | RunPod | Rent |
| RTX 3070 | 8 GB | $0.130community | RunPod | Rent |
| RTX 4090 | 24 GB | $0.134marketplace | Vast.ai | Rent |
Related models
| Model | Cheapest in $/1M | Cheapest out $/1M | Providers |
|---|---|---|---|
| Llama 3.3 70B | $0.100 | $0.200 | 15 |
| Llama 3.1 8B | $0.020 | $0.030 | 15 |
| Llama 4 Scout | $0.050 | $0.100 | 11 |
| Llama 3.1 70B Instruct | $0.120 | $0.300 | 10 |
| Llama 4 Maverick | $0.050 | $0.100 | 9 |
| Llama-3-70b | $0.120 | $0.300 | 7 |
FAQ
What is the cheapest API for Llama 3.2 3B?
As of 2026-08-23, the lowest input price for Llama 3.2 3B is DeepInfra at $0.020 per 1M input tokens. The lowest output price is DeepInfra at $0.020 per 1M output tokens. The cheapest provider for a real workload depends on its input/output mix. The next-lowest input price is Novita AI at $0.030, a 50% difference.
How much VRAM do you need to self-host Llama 3.2 3B?
Llama 3.2 3B has 3.2B parameters, so plan for roughly 5 GB of VRAM at FP8 or 3 GB at INT4/AWQ, KV-cache headroom included. A single V100 (32 GB) fits it.
← All models · GPU rental prices · Breakeven calculator · Get this data as JSON