Credits

Credits are spent per message — lighter models cost less, heavier models cost more. This is your plan's total budget.

Reset cadence

How often your credit balance refills back to the plan's limit.

Priority traffic

Requests are admitted from a reserved lane, so a busy period queues everyone else first.

Fast models

Gemini 3.1 Flash-Lite, Claude Haiku 4.5 and ChatGPT 5.4 Mini — one from each vendor, on every plan.

Reasoning and frontier models

ChatGPT 5.5 and 5.4 Reasoning, Gemini 3.1 Pro and 2.5 Pro — the models Auto reaches for on a hard question.

Claude (Sonnet family)

Claude Sonnet 5, 4.6, 4.5 and 3.7 from Anthropic.

Total models unlocked

How many of Evelant’s models you can pick from at any time.

Bring your own API key (BYOK)

Connect an Anthropic, OpenAI, Google, DeepSeek, Kimi, GLM, Qwen or Ollama account and run those models on your own credentials — they cost no credits.

Page context & smart summary

Evelant reads the page you’re on and can summarize it instantly.

Full-page deep context

Sends the entire page as context instead of only the most relevant excerpts.

Deep Scan for long documents & PDFs

Reads long documents and PDFs beyond the standard context window, not just the first portion.

Web search with verified sources

Automatically searches the web for fresh information, with cited sources.

Multi-tab sessions

Keep independent AI conversations running across every browser tab.

File & image attachments

Attach PDFs, screenshots, or images directly in chat. Needs an account — guest sessions cannot attach.

Shared rooms

Invite someone into a conversation and watch the same answer stream in for both of you.