Skip to content
NewTry RouterPlus with free creditsTry playground →
RouterPlus
Book a demo

Phala API

Open in consoleI need inference
  • 13Models
  • 6Labs
  • 1.05MMax context
  • $0.15/ $0.50From, per 1M in / out

Pricing

ModelLabContextInput / 1MCached / 1MOutput / 1MProviders
DeepSeek V4.1 Flash 1017deepseek/deepseek-v4.1-flash-1017DeepSeek1.05M$0.75$0.15$1.606
GLM 5.3 Flashz-ai/glm-5.3-flashZ.ai1.05M$0.15$0.03$0.5012
GLM 5.3z-ai/glm-5.3Z.ai1.05M$1.40$0.26$4.4012
Qwen3.8 27Bqwen/qwen3.8-27bQwen262K$0.20$0.05$2.508
DeepSeek V4 Flash 0731deepseek/deepseek-v4-flash-0731DeepSeek1.05M$0.44$0.028$1.329
Kimi K3moonshotai/kimi-k3Moonshot AI1.05M$3.00$0.30$15.0010
Qwen3.6 27Bqwen/qwen3.6-27bQwen262K$0.32$0.03$3.254
Qwen3.6 35B A3Bqwen/qwen3.6-35b-a3bQwen262K$0.20$0.056$1.274
Kimi K2.6moonshotai/kimi-k2.6Moonshot AI262K$1.09$0.37$4.605
GLM 5.1z-ai/glm-5.1Z.ai203K$1.21$0.60$4.202
DeepSeek V3.2deepseek/deepseek-v3.2DeepSeek131K$1.00$0.50$1.002
gpt-oss-120bopenai/gpt-oss-120bOpenAI131K$0.15–$0.602
Llama 3.3 70B Instructmeta-llama/llama-3.3-70b-instructMeta131K$2.00–$2.002

Phala's price, USD, as of November 5, 2026. Live prices

Quickstart

OpenAI-compatible
curl https://api.routerplus.com/v1/chat/completions \
  -H "Authorization: Bearer $ROUTERPLUS_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "z-ai/glm-5.3",
    "messages": [{ "role": "user", "content": "Hello" }],
    "max_tokens": 1024,
    "provider": { "only": ["phala"] }
  }'

FAQ

How do I send a request to Phala?

Name Phala in provider.only or provider.order, as "phala" or "phala-chat". Phala serves requests that name it. Where Phala is the last provider still serving a model, it also serves requests that do not name it.

How much does Phala cost on RouterPlus?

You pay Phala's own price for the model, which can be higher than the price where routing starts. Phala's lowest input price is $0.15 per million tokens (GLM 5.3 Flash) and its lowest output price is $0.50 per million tokens (GLM 5.3 Flash). Prices as of November 5, 2026.

Which models does Phala serve?

DeepSeek V4.1 Flash 1017, GLM 5.3 Flash, GLM 5.3, Qwen3.8 27B, DeepSeek V4 Flash 0731, Kimi K3, Qwen3.6 27B, Qwen3.6 35B A3B, Kimi K2.6, GLM 5.1, DeepSeek V3.2, gpt-oss-120b and Llama 3.3 70B Instruct.

What does the Phala enclave protect?

Your prompt passes through the RouterPlus gateway before it reaches Phala, so the enclave protects it at Phala, not on our side. Phala routes each model to one of several third-party TEE backends. RouterPlus neither chooses nor sees which one serves a call.

Does Phala keep my prompts?

Zero data retention is requested from Phala on every call. It is Phala's commitment, not something RouterPlus measures.

Can I verify Phala's attestation?

Phala publishes an attestation report and signs a receipt for each answer. RouterPlus does not check them per request.

Why did Phala return a 429?

Phala's capacity is smaller than other providers'. A long prompt, or one without max_tokens, can be refused with a 429 that waiting will not fix. Shorten the prompt, set max_tokens, or allow another provider.

Route to Phala with one key

Open in consoleI need inference