GPT vs Llama
The short answer. The cheapest GPT model we track, gpt-5-nano, costs $0.400 per million output tokens and $0.050 per million input tokens. The cheapest Llama model, Llama 3.1 8B Instruct, costs $0.080 per million output tokens and $0.050 per million input tokens, as sold by OpenRouter. At the entry level, Llama is 5.0 times cheaper than GPT on output tokens. Prices read on 3 November 2026 and 31 October 2026.
This page compares what OpenAI and Meta charge through their APIs. Every price below carries the day we read it and links to the page it came from. We do not rank quality: we tell you what each model costs, and since when.
At the top of each range
- Among GPT models that carry a relayed score, the highest on the Artificial Analysis intelligence index is GPT-6 Astra, at 56.9, sold by OpenRouter at $50.00 per million output tokens.
- Among Llama models that carry a relayed score, the highest on the Artificial Analysis intelligence index is Llama 4 Maverick, at 9.3, sold by OpenRouter at $0.6525 per million output tokens.
The index is produced by Artificial Analysis and relayed through OpenRouter's catalogue. It is not our measurement; we record it as a level 4 source and show it next to the price we read. A model without a score here has not been scored yet, which is not the same as scoring low.
Every model, side by side
GPT models
34 standard offers, from the prices OpenAI publishes itself.
| Model | Input | Output | Context | Read on | Source |
|---|---|---|---|---|---|
| gpt-5-nano | $0.050 | $0.400 | 31 October 2026 | OpenAI | |
| gpt-6-luna | $0.100 | $0.500 | 31 October 2026 | OpenAI | |
| gpt-4o-mini | $0.150 | $0.600 | 31 October 2026 | OpenAI | |
| gpt-5.6-luna | $0.200 | $1.20 | 31 October 2026 | OpenAI | |
| gpt-5.4-nano | $0.200 | $1.25 | 31 October 2026 | OpenAI | |
| gpt-4.1-mini | $0.400 | $1.60 | 31 October 2026 | OpenAI | |
| gpt-5-mini | $0.250 | $2.00 | 31 October 2026 | OpenAI | |
| o3-mini | $1.10 | $4.40 | 31 October 2026 | OpenAI | |
| o4-mini | $1.10 | $4.40 | 31 October 2026 | OpenAI | |
| gpt-5.4-mini | $0.750 | $4.50 | 31 October 2026 | OpenAI | |
| gpt-6.1-terra | $1.00 | $5.00 | 27 October 2026 | OpenAI | |
| gpt-4.1 | $2.00 | $8.00 | 31 October 2026 | OpenAI | |
| o3 | $2.00 | $8.00 | 31 October 2026 | OpenAI | |
| gpt-5 | $1.25 | $10.00 | 31 October 2026 | OpenAI | |
| gpt-5.1 | $1.25 | $10.00 | 31 October 2026 | OpenAI | |
| gpt-6-sol | $2.00 | $10.00 | 31 October 2026 | OpenAI | |
| gpt-6.1-sol | $2.00 | $10.00 | 31 October 2026 | OpenAI | |
| gpt-4o | $2.50 | $10.00 | 31 October 2026 | OpenAI | |
| gpt-5.6-terra | $2.00 | $12.00 | 31 October 2026 | OpenAI | |
| gpt-5.2 | $1.75 | $14.00 | 31 October 2026 | OpenAI | |
| gpt-5.4 (<272K context length) | $2.50 | $15.00 | 31 October 2026 | OpenAI | |
| gpt-4o-2024-05-13 | $5.00 | $15.00 | 31 October 2026 | OpenAI | |
| gpt-5.6-sol | $4.00 | $20.00 | 31 October 2026 | OpenAI | |
| gpt-5.5 (<272K context length) | $5.00 | $30.00 | 31 October 2026 | OpenAI | |
| gpt-4-turbo-2024-04-09 | $10.00 | $30.00 | 31 October 2026 | OpenAI | |
| gpt-6-astra | $10.00 | $50.00 | 31 October 2026 | OpenAI | |
| o1 | $15.00 | $60.00 | 31 October 2026 | OpenAI | |
| gpt-4-0613 | $30.00 | $60.00 | 31 October 2026 | OpenAI | |
| o3-pro | $20.00 | $80.00 | 31 October 2026 | OpenAI | |
| gpt-5-pro | $15.00 | $120 | 31 October 2026 | OpenAI | |
| gpt-5.2-pro | $21.00 | $168 | 31 October 2026 | OpenAI | |
| gpt-5.4-pro (<272K context length) | $30.00 | $180 | 31 October 2026 | OpenAI | |
| gpt-5.5-pro (<272K context length) | $30.00 | $180 | 31 October 2026 | OpenAI | |
| o1-pro | $150 | $600 | 31 October 2026 | OpenAI |
Llama models
8 standard offers, from OpenRouter's prices: we track no price list published by Meta itself for these models.
| Model | Input | Output | Context | Read on | Source |
|---|---|---|---|---|---|
| Llama 3.1 8B Instruct | $0.050 | $0.080 | 131 072 | 3 November 2026 | OpenRouter |
| Llama Guard 4 12B | $0.180 | $0.180 | 163 840 | 3 November 2026 | OpenRouter |
| Llama 3.2 1B Instruct | $0.027 | $0.201 | 60 000 | 31 October 2026 | OpenRouter |
| Llama 4 Scout | $0.100 | $0.300 | 1 310 720 | 3 November 2026 | OpenRouter |
| Llama 3.2 3B Instruct | $0.0509 | $0.335 | 131 072 | 3 November 2026 | OpenRouter |
| Llama 3.3 70B Instruct | $0.200 | $0.520 | 131 072 | 3 November 2026 | OpenRouter |
| Llama 4 Maverick | $0.1875 | $0.6525 | 524 288 | 3 November 2026 | OpenRouter |
| Llama 3.1 70B Instruct | $0.720 | $0.720 | 131 072 | 3 November 2026 | OpenRouter |
How these prices have moved
We hold dated captures of these models' prices back to 26 July 2023. We did not observe these values ourselves: they were recovered from public captures of OpenRouter's catalogue, and each links to its capture so you can check it without trusting us. Largest moves on output price:
| Model | Before | After | Change | Captured |
|---|---|---|---|---|
| Meta: Llama 3.2 1B Instruct | $0.010 | $0.200 | +1900% | 14 November 2025 |
| Meta: Llama 3.2 3B Instruct | $0.020 | $0.340 | +1600% | 3 March 2026 |
| Meta: Llama 3.2 3B Instruct | $0.006 | $0.024 | +300% | 4 September 2025 |
| Meta: Llama Guard 4 12B | $0.050 | $0.180 | +260% | 12 August 2025 |
| Meta: Llama 3.3 70B Instruct | $0.036 | $0.120 | +233% | 24 September 2025 |
| Meta: Llama 3.3 70B Instruct | $0.120 | $0.390 | +225% | 12 October 2025 |
Questions
Which is cheaper, GPT or Llama?
The cheapest GPT model we track, gpt-5-nano, costs $0.400 per million output tokens and $0.050 per million input tokens. The cheapest Llama model, Llama 3.1 8B Instruct, costs $0.080 per million output tokens and $0.050 per million input tokens, as sold by OpenRouter. At the entry level, Llama is 5.0 times cheaper than GPT on output tokens. Prices read on 3 November 2026 and 31 October 2026.
What is the cheapest GPT model?
gpt-5-nano, at $0.400 per million output tokens, read on 31 October 2026.
What is the cheapest Llama model?
Llama 3.1 8B Instruct, at $0.080 per million output tokens, read on 3 November 2026.
Where do these prices come from?
From the price lists OpenAI and OpenRouter publish, read by our collectors. No figure on this page was produced by a language model.