The models behind AI apps and agents. Each profile covers what the model is built for, its API prices per million tokens and where you can use it: in an app, through an API or both.
Newest first
Coding & development
OpenAI's mid-priced GPT-6.1 model for complex coding, computer use and professional work.
OpenAI API: $2 per million input and $10 per million output tokens, $0.10 cached input. In ChatGPT, available in Work and Codex on paid plans.
Productivity
OpenAI's lower-cost GPT-6.1 model for screen work, images of documents and everyday writing.
OpenAI API: $1 per million input and $5 per million output tokens, $0.05 cached input. In ChatGPT, available in Work and Codex on Plus, Pro, Enterprise and Edu.
AI agents
Anthropic's highest-priced Claude model for demanding reasoning and long-horizon agent work.
Claude API: $10 per million input and $50 per million output tokens. Not on Claude Free; on Pro via usage credits; on Max within 50% of weekly limits.
Coding & development
OpenAI's top GPT-6 model for long coding, computer-use and document tasks.
OpenAI API: $10 per million input and $50 per million output tokens; prompts over 272K input tokens cost 2x input and 1.5x output. In ChatGPT, included in Plus ($20 a month), Pro (from $100), Business and Enterprise.
Coding & development
Anthropic's mid-priced Claude model for everyday coding, documents and agent steps.
Claude API: $2 per million input and $10 per million output tokens; cache reads $0.10; Batch API 50% off. In the Claude apps, Sonnet is listed on Free, Pro ($20 a month, or $17 with annual billing) and Max (from $100).
Coding & development
Anthropic's low-cost API model for classifying, extracting and routing at volume.
Claude API: $0.10 per million input and $0.50 per million output tokens for prompts up to 100,000 tokens; $0.50 and $2.50 above that. Prompt caching is billed separately.
Coding & development
Moonshot AI's open-weight 3T-class model family for long coding and knowledge-work tasks.
Kimi API, pay as you go: K3.1 Pro $2 per million input and $10 per million output tokens; K3.1 Pro Highspeed $4 and $20; K3 $3 and $15. Kimi app: free Adagio plan, paid plans from $19 per month (Plus) to $199 per month (Ultra), lower with annual billing.
Coding & development
Evaluate code and screenshots together in an API workflow.
DeepSeek API (deepseek-flash, now the V4.1 Flash 1017 snapshot), per 1M tokens: $0.30 input, $0.006 cached input, $1.20 output at peak times; half that off-peak ($0.15, $0.003, $0.60). Other providers set their own rates. MIT-licensed weights.
Coding & development
Google's flagship reasoning model for long software, science and agent tasks.
Gemini API, paid tier only: $2 per million input tokens and $10 per million output tokens through November 30, 2026, then $4 and $20; Batch and Flex at half price. In the Gemini app for Google AI Pro and Ultra subscribers; subscription prices vary by country.
Coding & development
Explore a long code or document task with a multimodal model.
Paid API usage · $1.00 per million input tokens and $2.70 per million output tokens at the four providers listed on OpenRouter
Coding & development
Evaluate a text model for planned, multi-step coding work.
Paid API usage · About $0.75–$0.83 per million input tokens and $2.25–$2.50 per million output tokens, depending on provider and UTC hour
Coding & development
Build a visual coding trial around code, images and clear checks.
Paid API usage · Per-token rates vary by provider and configuration
Coding & development
Test a reasoning model on a difficult, text-based coding task.
Paid API usage · Per-token rates vary by provider and configuration
AI agents
Try text-based reasoning workflows on a free, rate-limited endpoint.
Free endpoint · Request limits and NVIDIA data-use conditions apply
Coding & development
Keep a specific text-model revision in a repeatable coding workflow.
Paid API usage · Per-token rates vary by provider and configuration
Coding & development
Bring mixed media into a model-assisted research or coding workflow.
Paid API usage · Per-token rates vary by provider and configuration
Coding & development
Evaluate complex code changes and dense source material through an API.
Paid API usage · Per-token rates vary by provider and configuration
Writing & editing
Draft, classify and structure repeated tasks through an API.
Paid API usage · Per-token rates vary by provider and configuration
A cheaper model that needs more attempts or longer reasoning can cost more per result. Compare input and output prices with your real prompt sizes.
Long coding sessions, document analysis and quick classification suit different models. Check context size, reasoning settings and tool support.
Some models are only in the API, others only in certain app plans. Confirm the access you need before you build around one.
Coding & development
Coding & development
Coding & development