cocodot pricing
cocodot has two products, priced independently: AI API relay (per token, never above official) and a US virtual card (open + per-use fees). Top up with Alipay — no overseas card needed.
→ Estimate your monthly API cost with the calculatorAI API relay
- · Claude / GPT / Gemini: up to 20% off official — most 10% off, live per-model prices below
- · Qwen & other Chinese models: much cheaper
- · Per-token billing, price capped at the model’s official rate
- · $0.5 free credit on email verification (API-only) — test the real model before topping up
US virtual card
- · Open a card: $9.9 one-time
- · Top-up fee: 3%
- · Spending: no separate spending fee; $0.60 settlement fee per purchase under $20 (from 2026-09-07; no settlement fee on purchases of $20 or more); from 2026-09-24, when the card issuer charges a fee on a transaction (such as a foreign transaction fee, 3DS verification fee or authorization fee), a corresponding fee is charged for that transaction — see your statement for the amount (shown as “Issuer transaction fee”)
- · Active monthly fee: $1
- · 100% of balance is spendable on your card
Top-up: Alipay (RMB) → USD balance on arrival. Test with a small amount first, then scale.
Per-model prices
Every model, every price tier, in one place — no sign-up needed. Billed per token, no monthly fee, no minimum, no top-up surcharge.
Claude · GPT · Gemini
Never above the model’s official price. Main models are 10% below it.
| Model | Model ID | Official (in/out) | Input / 1M | Output / 1M | You save | Input length |
|---|---|---|---|---|---|---|
| Claude Fable 5 | mcf-1 | $10.00 / $50.00 | $9.50 | $47.50 | -5% | flat rate |
| Claude Opus 4.8 | mco-6 | $5.00 / $25.00 | $4.50 | $22.50 | -10% | flat rate |
| Claude Opus 5 | mco-7 | $5.00 / $25.00 | $4.50 | $22.50 | -10% | flat rate |
| Claude Sonnet 5 | mcs-6 | $2.00 / $10.00 | $1.80 | $9.00 | -10% | flat rate |
| Claude Opus 4.7 | mco-5 | $5.00 / $25.00 | $4.50 | $22.50 | -10% | flat rate |
| Claude Sonnet 4.6 | mcs-5 | $3.00 / $15.00 | $2.40 | $12.00 | -20% | flat rate |
| Claude Haiku 4.5 | mch-1 | $1.00 / $5.00 | $0.90 | $4.50 | -10% | flat rate |
| GPT-5.5 | mog-6 | $5.00 / $30.00 | $4.50 $9.00 | $27.00 $40.50 | -10% | to 272k above |
| GPT-5.4 | mog-3 | $2.50 / $15.00 | $2.25 $4.50 | $13.50 $20.25 | -10% | to 272k above |
| GPT-5.6 Sol | mog-8-s | $4.00 / $20.00 | $3.60 $7.20 | $18.00 $27.00 | -10% | to 272k above |
| GPT-5.6 Terra | mog-8-t | $2.00 / $12.00 | $1.80 $3.60 | $10.80 $16.20 | -10% | to 272k above |
| GPT-5.6 Luna | mog-8-l | $0.20 / $1.20 | $0.18 $0.36 | $1.08 $1.62 | -10% | to 272k above |
| GPT-5.3-codex | mog-5 | $1.75 / $14.00 | $1.575 | $12.60 | -10% | flat rate |
| Gemini 3.1 Pro | mgg-8 | $2.00 / $12.00 | $1.80 $3.60 | $10.80 $16.20 | -10% | to 200k above |
| Gemini 3 Flash | mgg-7 | $0.50 / $3.00 | $0.45 | $2.70 | -10% | flat rate |
| Gemini 3.5 Flash | mgg-11 | $1.50 / $9.00 | $1.35 | $8.10 | -10% | flat rate |
| Gemini 3.6 Flash | mgg-12 | $1.50 / $7.50 | $1.35 | $6.75 | -10% | flat rate |
| Gemini 3.7 Flash | mgg-13 | $0.75 / $3.75 | $0.75 | $3.75 | = official | flat rate |
Chinese models
Priced against each vendor’s official rate, billed from your USD balance at the live FX rate.
| Model | Model ID | Official (in/out) | Input / 1M | Output / 1M | You save | Input length |
|---|---|---|---|---|---|---|
| Qwen3.5 Flash | qwen3.5-flash | $0.031 / $0.124 | $0.028 $0.084 $0.167 | $0.112 $0.335 $0.67 | -10% | to 128k to 256k above |
| Qwen3.5 35B-A3B | qwen3.5-35b-a3b | $0.062 / $0.496 | $0.056 $0.223 | $0.446 $1.785 | -10% | to 128k above |
| Qwen3.5 397B-A17B | qwen3.5-397b-a17b | $0.186 / $1.116 | $0.167 $0.418 | $1.004 $2.511 | -10% | to 128k above |
| Qwen3.6 Plus | qwen3.6-plus | $0.31 / $1.86 | $0.279 $1.116 | $1.674 $6.695 | -10% | to 256k above |
| GLM-5.1 | glm-5.1 | $0.93 / $3.72 | $0.883 $1.178 | $3.534 $4.123 | -5% | to 32k above |
| GLM-5.2 | glm-5.2 | $1.24 / $4.34 | $1.178 | $4.123 | -5% | flat rate |
| Qwen Flash | qwen-flash | $0.023 / $0.232 | $0.022 $0.088 $0.177 | $0.221 $0.883 $1.767 | -5% | to 128k to 256k above |
| GLM-5 | glm-5 | $0.62 / $2.79 | $0.558 $0.837 | $2.511 $3.069 | -10% | to 32k above |
| Kimi K3 | kimi-k3 | $3.10 / $15.498 | $3.10 | $15.498 | = official | flat rate |
| Qwen 3.8 Max | qwen3.8-max | $1.86 / $5.579 | $1.86 | $5.579 | = official | flat rate |
| GLM-5.3 | glm-5.3 | $1.24 / $4.34 | $1.24 | $4.34 | = official | flat rate |
| Qwen 3.7 Flash | qwen3.7-flash | $0.031 / $0.124 | $0.031 $0.093 $0.186 | $0.124 $0.372 $0.744 | = official | to 32k to 256k above |
| GPT-OSS 120B | gpt-oss-120b | $0.155 / $0.775 | $0.139 | $0.697 | -10% | flat rate |
| Qwen 3.7 Plus | qwen3.7-plus | $0.31 / $1.24 | $0.294 | $1.178 | -5% | flat rate |
| Qwen 3.7 Max | qwen3.7-max | $1.86 / $5.579 | $1.767 | $5.30 | -5% | flat rate |
| Qwen 3.6 Flash | qwen3.6-flash | $0.186 / $1.116 | $0.177 $0.707 | $1.06 $4.24 | -5% | to 128k above |
| Kimi K2.7 Code | kimi-k2.7-code | $1.007 / $4.185 | $0.957 | $3.975 | -5% | flat rate |
| Kimi K2.7 Code 高速版 | kimi-k2.7-code-highspeed | $2.015 / $8.369 | $1.914 | $7.951 | -5% | flat rate |
| 混元 T1 Vision | hunyuan-t1-vision-20250916 | $0.465 / $1.395 | $0.442 | $1.325 | -5% | flat rate |
| Kimi K2.6 | kimi-k2.6 | $1.007 / $4.185 | $0.957 | $3.975 | -5% | flat rate |
Tiered models: price depends on how long your input is. Every tier is listed above — there is no hidden surcharge for long context.
GLM-5.x and Gemini 3.6/3.7 Flash are reasoning models: they burn thinking tokens before the answer (measured: ~100 even for trivial prompts), and those count as output. Give max_tokens room (1000+), or you may get an empty answer. Qwen3.5 Flash, Qwen 3.7 Flash and GPT-OSS 120B are reasoning models too: they output their thinking first, and that also counts as output. With a small max_tokens you may get an empty answer, or the limit may be ignored and the model keeps going.
Prompt caching works here: cached input is billed at half the input rate, and usage returns the cache token counts so you can verify it yourself. Hit rate depends on how stable your prefix is — we do not promise a number.
Same prices via API: GET https://cocodot.co/api/ai/models
Pricing FAQ
How does cocodot charge?
Two independent prices: the API relay is billed per token (never above official); the virtual card has an open fee, a top-up fee and per-transaction fees incurred on spending. You only pay for what you use.
Is there a free quota?
Yes — verify your email to get $0.5 free API credit (usable only for API calls). Test the real model first, then top up small and scale up.
How do I top up?
Top up with Alipay (RMB). It converts to a USD balance on arrival. No overseas card needed.
Will the price ever exceed the official rate?
No. API usage is metered and capped at the model’s official price; premium models run 10–20% below official, per model.