⚠️ DeepSeek V4 / deepseek-chat is temporarily unavailable — our upstream account for this model is currently suspended, so requests to it return HTTP 503. The gateway and every other model keep working normally.
Available right now (the complete live set): kimi-k3 (256K context, flagship, $0.28/M cached input · $3.50/M fresh input · $18.00/M output), kimi-k2.7-code and kimi-k2.7-code-highspeed (coding), glm-4-flash (cheapest at $0.10/M in / $0.40/M out).
See the live model list | Try a working model in the playground | Get a free key

OpenAI-Compatible API

DeepSeek V4 API Access

Get OpenAI-compatible access to Chinese LLMs through a single endpoint. Note: DeepSeek V4 itself is paused right now (upstream billing suspended — deepseek-chat returns 503). The 4 models that actually serve traffic are Kimi K3, Kimi K2.7 Code, Kimi K2.7 Code HighSpeed and GLM-4 Flash. No complex setup. Start with $1 free credit.

Start Free Trial View API Docs

This model is currently paused. DeepSeek V4 is on our roster but cannot serve requests right now — its upstream billing is not settled, so calls return 503 model_unavailable. What actually serves traffic today: Kimi K3, Kimi K2.7 Code, Kimi K2.7 Code HighSpeed and GLM-4 Flash. Read the authoritative list at GET /v1/models, or get a free key and try a working model.

Why Use TokenEase for DeepSeek API?

⚡

Single Endpoint

Access DeepSeek V4, Kimi K3, GLM-5.1, Qwen-Plus, Doubao Pro, and Hunyuan through one OpenAI-compatible API. Switch models by changing one parameter.

🔒

Reliable Fallback

When DeepSeek is overloaded or unavailable, automatically route to equivalent models. No downtime, no code changes needed.

💰

40% Cheaper Than OpenRouter

Same OpenAI-compatible surface, lower pricing. Kimi K3, the live flagship, costs $0.28/M cached input · $3.50/M fresh input · $18.00/M output. GLM-4 Flash costs $0.10/M in / $0.40/M out. Transparent pricing, no hidden fees.

🚀

$1 Free Trial

Test all 4 live models with $1 free credit. No credit card required. See which model works best for your use case before committing.

🛠️

Drop-in Replacement

Works with any OpenAI SDK, LangChain, LlamaIndex, or custom code. Just change the base URL and API key. Same request/response format.

📊

Usage Analytics

Track token usage, costs, and model performance in real-time. Optimize your spend by choosing the right model for each task.

Model Pricing (per million tokens)

Model	Input	Output	Context	Status
DeepSeek V4	$0.50	$2.00	128K	High Demand
Kimi K3	$0.60	$2.40	256K	Available
GLM-5.1	$0.55	$2.20	128K	Available
Qwen-Plus	$0.40	$1.20	128K	Available
Doubao Pro	$0.35	$1.40	128K	Available
Hunyuan	$0.45	$1.80	128K	Available

TokenEase vs Direct DeepSeek API

OpenAI-compatible format TokenEase: Yes

Multiple model fallback TokenEase: Yes (4 live models)

Single API key for all models TokenEase: Yes

Usage analytics dashboard TokenEase: Yes

Free trial TokenEase: $1 credit

Credit card required to try TokenEase: No

5-Minute Integration

                # Python - just change base_url and api_key
from openai import OpenAI

client = OpenAI(
    base_url="https://tokenease.io/v1",
    api_key="your-tokenease-api-key"
)

# Use Kimi K3 (DeepSeek V4 is temporarily unavailable)
response = client.chat.completions.create(
    model="kimi-k3",
    messages=[{"role": "user", "content": "Hello!"}]
)

# Or instantly switch to Kimi K2.7 Code, GLM-4 Flash, etc.
response = client.chat.completions.create(
    model="kimi-k3",
    messages=[{"role": "user", "content": "Hello!"}]
)
            

                # cURL
curl https://tokenease.io/v1/chat/completions \\
  -H "Authorization: Bearer your-tokenease-api-key" \\
  -H "Content-Type: application/json" \\
  -d '{"model": "kimi-k3", "messages": [{"role": "user", "content": "Hello!"}]}'
            