Gemini API Without a Google Account: 3 Flash / 3.7 Flash / 3.1 Pro Pricing (2026)
Google's own API needs an overseas card and blocks some regions. cocodot serves Gemini through an OpenAI-compatible endpoint, Alipay or card — Gemini 3 Flash from $0.45/$2.7 per million tokens, 3.1 Pro and 3.5/3.6 Flash at 90% of list, 3.7 Flash at list price.
The two walls
Google AI Studio billing wants an overseas card; the API itself is unavailable in several regions. cocodot's answer is an official-relay, OpenAI-compatible endpoint: no Google account, no card gymnastics — change base_url and call.
How to split the models
3 Flash: high-volume, latency-sensitive, low-value-per-call work (labeling, cleaning, first-pass screening), plus the daily driver role — chatbots, summarization, RAG generation. 3.6/3.7 Flash: newer reasoning Flash models with 1M context; they spend thinking tokens before answering (billed as output, so set max_tokens 1000+) — for work that needs some reasoning without Pro pricing. 3.5 Flash: multimodal input with 1M context. Pro: cross-document reasoning, complex agents, long-context analysis (mind the >200K tier, fully listed above).
Two-minute setup
OpenAI-compatible: base_url https://cocodot.co/api/ai/v1, model set to the code — mgg-7 / mgg-13 / mgg-8 and the rest listed above (using the code directly is recommended). Any OpenAI SDK works — two config lines to migrate an existing project.
Gemini models on cocodot (USD per M tokens; 3.7 Flash at list price, the rest 90% of list)
| Model | ID | In / Out | Role |
|---|---|---|---|
| Gemini 3 Flash | mgg-7 | $0.45 / $2.7 | High-speed general: chat, summaries, realtime |
| Gemini 3.5 Flash | mgg-11 | $1.35 / $8.1 | Multimodal Flash: 1M context, image input |
| Gemini 3.6 Flash | mgg-12 | $1.35 / $6.75 | Newer reasoning Flash: 1M context, thinks first |
| Gemini 3.7 Flash | mgg-13 | $0.75 / $3.75 (= list price) | Latest Flash: 1M context, thinks first |
| Gemini 3.1 Pro | mgg-8 | $1.8 / $10.8 (>200K: $3.6/$16.2) | Complex reasoning, long context |