Compare API pricing for 25+ AI models including GPT-4o, Claude 3.5 Sonnet, Gemini 1.5 Pro, Llama 3, Flux, and more. Filter by model type, provider, and sort by price to find the best model for your budget.
| Model | Provider | Type | Input /1K tokens | Output /1K tokens | Per Image | Context | Notes |
|---|---|---|---|---|---|---|---|
| Gemini 1.5 Flash | Text | $0.00007 | $0.0003 | — | 1000K | Free tier available | |
| Gemini 2.0 Flash | Text | $0.00010 | $0.0004 | — | 1000K | Fast and affordable | |
| GPT-4o mini | OpenAI | Text | $0.00015 | $0.0006 | — | 128K | Fast & affordable |
| Mistral Small | Mistral AI | Text | $0.00020 | $0.0006 | — | 128K | Cost-efficient |
| Claude 3.5 Haiku | Anthropic | Text | $0.00080 | $0.0040 | — | 200K | Fastest & cheapest |
| Llama 3.1 70B | Meta (via providers) | Text | $0.00090 | $0.0009 | — | 128K | Open weights |
| o3-mini | OpenAI | Reasoning | $0.00110 | $0.0044 | — | 200K | Latest efficient reasoning |
| Gemini 1.5 Pro | Text | $0.00125 | $0.0050 | — | 2000K | 2M token context | |
| Gemini 2.5 Pro | Text | $0.00125 | $0.0100 | — | 1000K | Latest reasoning model | |
| GPT-4o | OpenAI | Text | $0.00250 | $0.0100 | — | 128K | Flagship multimodal model |
| Llama 3.1 405B | Meta (via providers) | Text | $0.00270 | $0.0027 | — | 128K | Open weights, self-host free |
| o1-mini | OpenAI | Reasoning | $0.00300 | $0.0120 | — | 128K | Affordable reasoning |
| Claude 3.5 Sonnet | Anthropic | Text | $0.00300 | $0.0150 | — | 200K | Best intelligence/value |
| Claude 4 Sonnet | Anthropic | Text | $0.00300 | $0.0150 | — | 200K | Latest generation |
| Mistral Large 2 | Mistral AI | Text | $0.00300 | $0.0090 | — | 128K | Strong European LLM |
| GPT-4 Turbo | OpenAI | Text | $0.01000 | $0.0300 | — | 128K | Previous gen flagship |
| o1 | OpenAI | Reasoning | $0.01500 | $0.0600 | — | 200K | Advanced reasoning |
| Claude 3 Opus | Anthropic | Text | $0.01500 | $0.0750 | — | 200K | Most powerful Claude 3 |
| Ideogram 2.0 Turbo | Ideogram | Image | — | — | $0.025 | — | Fast, affordable |
| Stable Image Core | Stability AI | Image | — | — | $0.030 | — | Affordable quality |
| DALL-E 3 (Standard) | OpenAI | Image | — | — | $0.040 | — | 1024×1024 per image |
| Flux 1.1 Pro | Black Forest Labs | Image | — | — | $0.040 | — | Top open-source image model |
| DALL-E 3 (HD) | OpenAI | Image | — | — | $0.080 | — | Higher quality |
| Stable Image Ultra | Stability AI | Image | — | — | $0.080 | — | Highest quality |
| Ideogram 2.0 | Ideogram | Image | — | — | $0.080 | — | Excellent text rendering |
Prices updated periodically. Always verify with official provider pricing pages. Prices shown per 1,000 tokens ($/1K) for text models.
Low-volume tasks (less than 10K requests/month): any model works. High-volume tasks (1M+ requests/month): even small cost differences multiply significantly. Calculate your expected token usage using our AI Token Calculator.
Simple Q&A and text formatting → GPT-4o mini or Gemini Flash. Complex reasoning and analysis → GPT-4o, Claude 3.5 Sonnet, or Gemini 1.5 Pro. Multi-file coding tasks → Claude 3.5 Sonnet (200K context). Long documents → Gemini 1.5 Pro (2M context).
Use our GPT Cost Calculator or AI Pricing Comparison tool. Enter your expected requests per month, average input tokens (your prompts), and average output tokens (expected responses) to get an exact monthly estimate.
OpenAI and Anthropic offer prompt caching which dramatically reduces costs for repeated system prompts. GPT-4o mini with caching can be 10x cheaper than GPT-4o for high-volume applications with consistent system prompts.
GPT-4o costs $0.0025 per 1,000 input tokens and $0.010 per 1,000 output tokens as of 2025. GPT-4o mini is significantly cheaper at $0.00015/1K input and $0.0006/1K output.
Claude 3.5 Sonnet costs $0.003/1K input and $0.015/1K output. GPT-4o costs $0.0025/1K input and $0.010/1K output. For most use cases, GPT-4o is slightly cheaper, but Claude offers a 200K context window vs GPT-4o's 128K.
GPT-4o mini ($0.00015/1K input) and Gemini 2.0 Flash ($0.0001/1K input) offer the best quality-to-cost ratio. Both are significantly cheaper than flagship models while maintaining strong performance on most tasks.
DALL-E 3 costs $0.040-0.120/image. Stable Image Core costs $0.030/image. Flux 1.1 Pro costs around $0.040/image. Ideogram 2.0 Turbo is $0.025/image. Open-source models like Stable Diffusion can be self-hosted for near-zero marginal cost.