Model Comparison
Pricing updated daily · Last refreshed Oct 15, 2026Context windows, output limits and per-million-token pricing across providers. Sort any column; filter to the providers you actually ship on.
| Model | Provider | Context window | Input $/1M ↑ | Output $/1M | Max output | Multimodal |
|---|---|---|---|---|---|---|
| Llama 3.2 1B Instruct | Meta | 60,000 | $0.03 | $0.20 | 54,000 | |
| Command R7B (12-2024) | Cohere | 128,000 | $0.04 | $0.15 | 4,000 | |
| Claude Haiku 5.5 | Anthropic | 1,000,000 | $0.10 | $0.50 | 128,000 | |
| GPT-6 Luna | OpenAI | 1,050,000 | $0.10 | $0.50 | 128,000 | |
| GPT-6 Luna Pro | OpenAI | 1,050,000 | $0.10 | $0.50 | 128,000 | |
| Command R (08-2024) | Cohere | 128,000 | $0.15 | $0.60 | 4,000 | |
| Mistral Small 4 | Mistral | 262,144 | $0.15 | $0.60 | 209,715 | |
| Llama 4 Scout | Meta | 1,310,720 | $0.18 | $0.59 | 117,964 | |
| Llama Guard 4 12B | Meta | 163,840 | $0.18 | $0.18 | 16,384 | |
| Llama 4 Maverick | Meta | 1,048,576 | $0.19 | $0.65 | 16,384 | |
| Ministral 3 14B 2512 | Mistral | 262,144 | $0.20 | $0.20 | 209,715 | |
| Command A+ | Cohere | 192,000 | $0.30 | $1.50 | 64,000 | |
| Devstral 2 2512 | Mistral | 262,144 | $0.40 | $2.00 | 209,715 | |
| Llama 3.3 70B Instruct | Meta | 131,072 | $0.59 | $0.79 | 32,768 | |
| Mistral Large 4 | Mistral | 1,048,576 | $0.68 | $2.09 | 262,144 | |
| Gemini 3.6 Flash | 1,048,576 | $0.75 | $3.75 | 65,536 | ||
| Gemini 3.7 Flash | 1,048,576 | $0.75 | $3.75 | 65,536 | ||
| Gemini 3.8 Flash | 1,048,576 | $0.75 | $3.75 | 65,536 | ||
| Nano Banana 2.1 | 65,536 | $1.50 | $7.50 | 58,982 | ||
| Mistral Medium 3.5 | Mistral | 262,144 | $1.50 | $7.50 | 209,715 | |
| Claude Sonnet 5.5 | Anthropic | 1,000,000 | $2.00 | $10.00 | 128,000 | |
| Gemini 4 Argon | 1,048,576 | $2.00 | $10.00 | 65,536 | ||
| GPT-6 Sol Pro | OpenAI | 1,050,000 | $2.00 | $10.00 | 128,000 | |
| GPT-6.1 Sol | OpenAI | 1,050,000 | $2.00 | $10.00 | 128,000 | |
| GPT-6.1 Sol Pro | OpenAI | 1,050,000 | $2.00 | $10.00 | 128,000 | |
| Command A | Cohere | 256,000 | $2.50 | $10.00 | 8,192 | |
| Command R+ (08-2024) | Cohere | 128,000 | $2.50 | $10.00 | 4,000 | |
| Claude Opus 5.5 | Anthropic | 1,000,000 | $4.00 | $20.00 | 128,000 | |
| Claude Opus 5 | Anthropic | 1,000,000 | $5.00 | $25.00 | 128,000 | |
| Claude Fable 5.1 | Anthropic | 1,000,000 | $10.00 | $50.00 | 128,000 |
Showing 30 of 30 models.
Where these numbers come from
Pricing is read from OpenRouter’s public model catalogue once a day. OpenRouter is a routing marketplace, so for a few models its rate can differ from a provider’s own list price — check the vendor’s pricing page before signing a contract on these figures.
Context sensitivity
A large context window is not the same as usable recall. Retrieval accuracy typically degrades well before the advertised limit, so treat the context column as a ceiling rather than a promise.
Frequently asked questions
Where does the model data come from?
Context windows, output limits and prices come from OpenRouter’s public model catalogue, read once a day. If the catalogue is unavailable the page shows a committed snapshot and says how old it is.
Why can a price here differ from the provider’s own pricing page?
OpenRouter is a routing marketplace, and for some models its listed rate differs from the provider’s direct list price. Treat this table as a comparison aid and confirm the figure on the vendor’s pricing page before you commit to a budget.
What is the difference between context window and max output?
The context window is the total number of tokens a model can handle in a single request, covering your input and its reply together. Max output is the cap on how many tokens it can generate in one response. A large window is a ceiling, not a promise: retrieval accuracy typically degrades well before the advertised limit.
Can I sort, filter or export the table?
Yes. You can sort by any column, search by name, filter by provider, show only multimodal models, and export the filtered view as a CSV file. All of it happens in your browser.