The complete OpenAI model lineup: 61 models spanning the GPT conversational series, o-series reasoning models, and budget-friendly mini variants. Compare scores, API pricing, context windows, and capabilities side by side.
| Model | Score | Input $/1M | Output $/1M |
|---|---|---|---|
| gpt-oss-120b (free) | 56 | Free | Free |
| gpt-oss-20b (free) | 56 | Free | Free |
| Model | Score | Input $/1M | Output $/1M |
|---|---|---|---|
| gpt-oss-20b | 46 | $0.030 | $0.140 |
| gpt-oss-120b | 46 | $0.039 | $0.190 |
| gpt-oss-120b (exacto) | 46 | $0.039 | $0.190 |
| gpt-oss-safeguard-20b | 53 | $0.075 | $0.300 |
| GPT-5 Nano | 64 | $0.050 | $0.400 |
| GPT-4.1 Nano | 58 | $0.100 | $0.400 |
| GPT-4o-mini (2024-07-18) | 44 | $0.150 | $0.600 |
| GPT-4o-mini | 44 | $0.150 | $0.600 |
| GPT-4o-mini Search Preview | 44 | $0.150 | $0.600 |
| Model | Score | Input $/1M | Output $/1M |
|---|---|---|---|
| GPT-3.5 Turbo | 31 | $0.500 | $1.50 |
| GPT-4.1 Mini | 58 | $0.400 | $1.60 |
| GPT-5 Mini | 65 | $0.250 | $2.00 |
| GPT-5.1-Codex-Mini | 60 | $0.250 | $2.00 |
| GPT-3.5 Turbo (older v0613) | 29 | $1.00 | $2.00 |
| GPT-3.5 Turbo Instruct | 26 | $1.50 | $2.00 |
| GPT-5 Image Mini | — | $2.50 | $2.00 |
| GPT Audio Mini | 49 | $0.600 | $2.40 |
| GPT-3.5 Turbo 16k | 31 | $3.00 | $4.00 |
| o4 Mini High | 61 | $1.10 | $4.40 |
| o4 Mini | 61 | $1.10 | $4.40 |
| o3 Mini High | 47 | $1.10 | $4.40 |
| o3 Mini | 47 | $1.10 | $4.40 |
| Model | Score | Input $/1M | Output $/1M |
|---|---|---|---|
| o4 Mini Deep Research | 66 | $2.00 | $8.00 |
| o3 | 62 | $2.00 | $8.00 |
| GPT-4.1 | 59 | $2.00 | $8.00 |
| GPT-5.1 | 67 | $1.25 | $10.00 |
| GPT-5 | 67 | $1.25 | $10.00 |
| GPT-5.1-Codex-Max | 66 | $1.25 | $10.00 |
| GPT-5.1-Codex | 62 | $1.25 | $10.00 |
| GPT-5 Codex | 62 | $1.25 | $10.00 |
| GPT-5.1 Chat | 61 | $1.25 | $10.00 |
| GPT-5 Chat | 57 | $1.25 | $10.00 |
| GPT-4o Audio | 53 | $2.50 | $10.00 |
| GPT Audio | 51 | $2.50 | $10.00 |
| GPT-4o (2024-11-20) | 50 | $2.50 | $10.00 |
| GPT-4o (2024-08-06) | 47 | $2.50 | $10.00 |
| GPT-4o Search Preview | 46 | $2.50 | $10.00 |
| GPT-4o | 45 | $2.50 | $10.00 |
| GPT-5 Image | — | $10.00 | $10.00 |
| GPT-5.2 | 68 | $1.75 | $14.00 |
| GPT-5.3-Codex | 67 | $1.75 | $14.00 |
| GPT-5.2-Codex | 67 | $1.75 | $14.00 |
| GPT-5.3 Chat | 62 | $1.75 | $14.00 |
| GPT-5.2 Chat | 62 | $1.75 | $14.00 |
| Model | Score | Input $/1M | Output $/1M |
|---|---|---|---|
| GPT-5.4 | 70 | $2.50 | $15.00 |
| GPT-4o (2024-05-13) | 45 | $5.00 | $15.00 |
| GPT-4o (extended) | 48 | $6.00 | $18.00 |
| GPT-4 Turbo | 46 | $10.00 | $30.00 |
| GPT-4 Turbo Preview | 40 | $10.00 | $30.00 |
| GPT-4 Turbo (older v1106) | 40 | $10.00 | $30.00 |
| o3 Deep Research | 74 | $10.00 | $40.00 |
| o1 | 65 | $15.00 | $60.00 |
| GPT-4 (older v0314) | 44 | $30.00 | $60.00 |
| GPT-4 | 44 | $30.00 | $60.00 |
| o3 Pro | 82 | $20.00 | $80.00 |
| GPT-5 Pro | 90 | $15.00 | $120.00 |
| GPT-5.2 Pro | 90 | $21.00 | $168.00 |
| GPT-5.4 Pro | 91 | $30.00 | $180.00 |
| o1-pro | 77 | $150.00 | $600.00 |
The GPT-4 series represents OpenAI's flagship conversational models. GPT-4o ("omni") is the multimodal variant that handles text, images, and audio natively. GPT-4.1 brings improved instruction following and coding performance. These models offer the best balance of capability and cost for general-purpose tasks.
The o1 and o3 models use chain-of-thought reasoning to "think" before responding. They excel at math, science, complex coding, and multi-step logic problems. o3-pro offers the highest reasoning quality at a premium price, while o3-mini provides reasoning capabilities at lower cost.
GPT-4o Mini is OpenAI's cost-optimized model, delivering strong performance at a fraction of the price of full GPT-4o. It supports vision, function calling, and JSON mode, making it ideal for high-volume production workloads where cost efficiency matters more than peak capability.
All OpenAI models are accessible through the OpenAI API and via aggregators like OpenRouter. Rate limits scale with usage tier (free, tier 1--5). Enterprise customers get higher throughput. Batch API offers 50% discounts for non-urgent workloads processed within a 24-hour window.