Call Portal M2-25 or Portal M3-55 and pay per token in USD.
A frontier agentic coding model. Built for long-context reasoning, precise edits, and tool use - exposed through a single stable model id.
Every call is metered per token in USD and billed against your wallet - monthly plan credit first, then your prepaid balance. See your real spend per call, with full token breakdowns, in Usage and Logs.
Priority tier - same capabilities as Portal M2-25 Non-fast, higher throughput and on-demand billing.
Top-tier general-purpose frontier model at Extra-High reasoning effort with a 1M-token context window. For complex agents, analysis and tool-heavy workflows.
Reasoning effort and context length don't change the per-token rate. Deeper reasoning spends more tokens, so hard problems cost more in total, never more per token. Metered per token in USD and billed against your wallet.
Flagship-speed frontier tier: near-top reasoning at priority speed with a 500K-class context window. The default when you want strong and fast at once.
This catalog contains Portal-branded customer routes. Underlying routing and availability may change.
Mint a key in the dashboard and point any OpenAI-compatible client at https://api.portal.ai/v1.