Vendor comparison
Meta vs Microsoft
Every Meta and Microsoft LLM model side by side: pricing per million tokens, context windows, and capabilities. Refreshed nightly from upstream.
Meta
14 models| Model | Context | In $/1M | Out $/1M |
|---|---|---|---|
| Llama 3 70B Instruct | 8k | 0.5100 | 0.7400 |
| Llama 3 8B Instruct | 8k | 0.0400 | 0.0400 |
|
Llama 3.1 70B Instruct
tools, json_mode
|
131k | 0.4000 | 0.4000 |
|
Llama 3.1 8B Instruct
tools, json_mode
|
16k | 0.0200 | 0.0500 |
|
Llama 3.2 11B Vision Instruct
vision, json_mode
|
131k | 0.2450 | 0.2450 |
| Llama 3.2 1B Instruct | 60k | 0.0270 | 0.2000 |
| Llama 3.2 3B Instruct | 80k | 0.0510 | 0.3400 |
| Llama 3.2 3B Instruct (free) | 131k | 0.0000 | 0.0000 |
|
Llama 3.3 70B Instruct
tools, json_mode
|
131k | 0.1000 | 0.3200 |
|
Llama 3.3 70B Instruct (free)
tools
|
66k | 0.0000 | 0.0000 |
|
Llama 4 Maverick
vision, json_mode
|
1049k | 0.1500 | 0.6000 |
|
Llama 4 Scout
vision, tools, json_mode
|
328k | 0.0800 | 0.3000 |
| Llama Guard 3 8B | 131k | 0.4800 | 0.0300 |
|
Llama Guard 4 12B
vision, json_mode
|
164k | 0.1800 | 0.1800 |
Microsoft
3 models| Model | Context | In $/1M | Out $/1M |
|---|---|---|---|
|
Phi 4
json_mode
|
16k | 0.0650 | 0.1400 |
|
Phi 4 Mini Instruct
json_mode
|
128k | 0.0800 | 0.3500 |
| WizardLM-2 8x22B | 66k | 0.6200 | 0.6200 |