Pricing
Pay only for the tokens you use. No monthly subscription fees. All prices include SeaLink's 4% platform fee.
Tiers
Same API. Model access and support level vary by tier.
Free
For trying the platform and shipping first experiments.
$1 trial credit on signup
Models included
Standard chat models (Haiku, GPT-4o mini, Qwen Turbo, Doubao Lite)
- $1 trial credit on signup
- Pay-as-you-go after credit is used
- Works with popular clients and integrations
- Community support
Detailed billing records — Not included
Pay-as-you-go
For teams shipping production workloads. No monthly fee — pay only for tokens.
Pay per token at catalog rates
Models included
All models — chat, reasoning, vision, embeddings, and audio
- All models, no restrictions
- Detailed billing records for reconciliation
- Budget alerts and per-key spending caps
- Multiple API keys for environments and teammates
- Email support
Detailed billing records — Included
Enterprise
For teams that need procurement support, SLA, and dedicated onboarding.
Models included
All models + priority routing for flagship and high-context models
- Everything in Pay-as-you-go
- Priority routing for flagship models
- Priority email support
- Custom invoice headers for procurement
- Uptime SLA — see /legal/sla
Detailed billing records — Included
Per-model rates
SeaLink adds a 4% platform fee on top of provider pricing — covering routing, unified billing, and billing records.
| Model | Provider | Input / 1M tokens | Output / 1M tokens | Context |
|---|---|---|---|---|
| GLM-4 Flash | Zhipu (GLM) | $0.00/1M | $0.00/1M | 128K |
| Cohere Rerank 3 | Cohere | $2.08/1M | $0.00/1M | 4K |
| Jina Reranker v2 | Jina AI | $0.52/1M | $0.00/1M | 8K |
| OpenAI Embedding 3 Large | OpenAI | $0.14/1M | $0.00/1M | 8K |
| Doubao 1.5 Lite 32k | ByteDance (Doubao) | $0.04/1M | $0.09/1M | 128K |
| Doubao 2.0 Mini | ByteDance (Doubao) | $0.04/1M | $0.09/1M | 128K |
| Qwen3 Turbo | Alibaba (Qwen) | $0.05/1M | $0.20/1M | 128K |
| Doubao 1.5 Pro 32k | ByteDance (Doubao) | $0.12/1M | $0.29/1M | 256K |
| DeepSeek V4 Flash | DeepSeek | $0.15/1M | $0.29/1M | 1000K |
| Mistral Small | Mistral AI | $0.10/1M | $0.31/1M | 128K |
| GLM-4.6 | Zhipu (GLM) | $0.14/1M | $0.43/1M | 200K |
| GPT-4o mini | OpenAI | $0.16/1M | $0.62/1M | 128K |
| Qwen3 Plus | Alibaba (Qwen) | $0.27/1M | $0.81/1M | 128K |
| Llama 3.3 70B Instruct | Meta (Llama) | $0.92/1M | $0.92/1M | 128K |
| Gemini 3.1 Flash-Lite | Google DeepMind (Gemini) | $0.26/1M | $1.56/1M | 1000K |
| Omni Moderation | OpenAI | $2.00/1M | $2.00/1M | 0K |
| Kimi K2 | Moonshot (Kimi) | $0.57/1M | $2.29/1M | 1000K |
| DeepSeek V4 Pro | DeepSeek | $1.81/1M | $3.62/1M | 1000K |
| Qwen3 Max | Alibaba (Qwen) | $0.81/1M | $4.06/1M | 128K |
| o3-mini | OpenAI | $1.14/1M | $4.58/1M | 200K |
| Gemini 3 Flash | Google DeepMind (Gemini) | $0.78/1M | $4.68/1M | 1000K |
| Claude Haiku 4.5 | Anthropic | $1.04/1M | $5.20/1M | 200K |
| Whisper large-v3 | OpenAI | $6.00/1M | $6.00/1M | 0K |
| Mistral Large | Mistral AI | $2.08/1M | $6.24/1M | 128K |
| GPT-4o | OpenAI | $2.60/1M | $10.40/1M | 128K |
| Gemini 3.1 Pro | Google DeepMind (Gemini) | $2.08/1M | $12.48/1M | 1000K |
| OpenAI TTS-1 HD | OpenAI | $15.00/1M | $15.00/1M | 0K |
| Claude Sonnet 4.6 | Anthropic | $3.12/1M | $15.60/1M | 200K |
| Claude Opus 4.7 | Anthropic | $5.20/1M | $26.00/1M | 1000K |
| Stable Diffusion 3.5 | Stability AI | $30.00/1M | $30.00/1M | 0K |
| ElevenLabs Multilingual v2 | ElevenLabs | $30.00/1M | $30.00/1M | 0K |
| DALL-E 3 | OpenAI | $40.00/1M | $40.00/1M | 0K |
| FLUX 1.1 Pro | Black Forest Labs (Flux) | $40.00/1M | $40.00/1M | 0K |
| Kling 1.6 | Kuaishou (Kling) | $120.00/1M | $120.00/1M | 0K |
| Runway Gen-4 | Runway | $150.00/1M | $150.00/1M | 0K |
| GPT-5.5 Pro | OpenAI | $31.20/1M | $187.20/1M | 272K |
All prices are per 1 million tokens, in USD. Local currency (THB / IDR / MYR) displayed at daily rates; settlement in USD.
What does the 4% fee cover?
SeaLink charges a 4% platform fee on every request. It's not just a markup — here's what you get for it:
Smart routing
Requests are routed to the best-performing upstream in real time. When one provider degrades, we fall over to the next without you changing a line of code.
One bill, 10+ providers
No need to sign contracts, reconcile invoices, or manage API keys with every model provider. One Singapore invoice covers everything — built for procurement.
Billing records & transparency
Per-request billing records show exactly which model served you, how many tokens you used, and what you paid. No black boxes.
SEA latency optimization
Singapore-hosted infrastructure with continuously tuned routing paths. Target ~150-250ms for chat models in SEA — verified by synthetic probes every 10 minutes.
Probe monitoring & status
Every model is continuously probed from Singapore. Live status, incident history, and uptime data are public at /status. Verify before you buy.
Procurement-ready materials
DPA, security questionnaires, entity verification, and invoice samples available on request. Legal and finance teams can evaluate SeaLink without chasing internal sales.
Compare: signing directly with 10 providers means managing 10 contracts, 10 invoices, 10 API keys, 10 billing rules, and 10 status pages. 4% is the price of not dealing with that.
What will it cost me?
Pick a common task, pick a model, slide the volume — see your monthly bill.
~ 800 in + 250 out / call
Estimated monthly cost
$21.32
≈ ¥154 · ฿768 · at today's rate
Estimates only. Actual usage billed per real tokens; unused balance refundable.
FAQ
Is there a monthly subscription fee?
No. SeaLink charges per token at catalog rates with a flat 4% platform fee. There are no recurring monthly charges, minimum commitments, or hidden fees.
Can you issue invoices?
Yes. Pay-as-you-go and Enterprise accounts receive monthly Singapore billing records with custom headers and tax IDs. Invoice and tax details appear when applicable.
Which payment methods?
Visa / Mastercard / Amex via Stripe. GrabPay, PromptPay, TrueMoney, GoPay, and GCash are available through local wallet payment links.
Can I set spending limits?
Yes. Each API key can have an optional monthly budget cap. You'll get alerts before hitting the limit.
What's the difference between Free and Pay-as-you-go?
Free includes $1 trial credit and access to standard chat models. Pay-as-you-go gives you all models, detailed billing records, budget alerts, multiple API keys, and email support — still no monthly fee. Enterprise adds priority routing, SLA, and custom invoice support.
Why does Claude Opus 4.7 sometimes cost more in practice than Opus 4.6 at the same per-token rate?
Anthropic shipped a new tokenizer with Opus 4.7. The same English text can encode into up to ~35% more tokens than Opus 4.6. Per-token price is unchanged; effective cost per request can rise. SeaLink passes through Anthropic's billing 1:1 plus our flat 4% — we don't add a markup on the tokenizer change.