SiliconFlow free models: limits, keys and failover

What is free on SiliconFlow

The platforms differ. The international pricing page lists no free models and offers $1 of starting credit. The China platform marks some models free, bills free-model calls at zero, and unlocks all free models after real-name verification.

Examples listed as free when we checked: tencent/Hunyuan-MT-7B , PaddlePaddle/PaddleOCR-VL-1.5 , BAAI/bge-m3 . Check the current catalog and your account access. Ficelle refreshes catalog data and applies eligibility checks during selection.

SiliconFlow free-tier limits

Both platforms say free-model rate limits are fixed but do not publish the numbers; per-model limits appear in the model marketplace. Over a limit, the API returns a 429.

Get a SiliconFlow API key

International keys are created at cloud.siliconflow.com. China-platform keys are created on cloud.siliconflow.cn, where real-name verification is required to use all free models.

https://cloud.siliconflow.com/account/ak

OpenAI compatibility

SiliconFlow supports most OpenAI parameters through the OpenAI SDK. The China platform uses https://api.siliconflow.cn/v1 and a separate account.

https://api.siliconflow.com/v1

Data and privacy

The international terms say SiliconFlow will not access, use or disclose your interaction data without authorization unless the law requires it.

Through Ficelle, your SiliconFlow key stays on your machine and prompts go straight to SiliconFlow: there is no Ficelle proxy in between.

Before you rely on it

• Match the account, API key and base URL to the same platform. A free listing on .cn does not establish free access through a .com key; rate limits are per account and per model.
• Free models are listed only on the China platform; the international platform gives $1 of credit instead.
• Several free models are specialised (translation, OCR, embeddings), not general chat models.

Use SiliconFlow with automatic failover

Ficelle can include SiliconFlow behind your local OpenAI-compatible endpoint when the integration is enabled and its models pass free-access and capability checks. SiliconFlow is listed in Ficelle Pro's maintained provider pack. Check your installed configuration, credentials and account entitlement before relying on it. Another eligible candidate must be available for failover to succeed. Local routing guards do not control upstream billing when an account permits paid usage beyond its free allocation. See strict-zero routing and account limits.

# SiliconFlow ships in Ficelle Pro's maintained provider pack.
ficelle doctor --text          # check this install before configuring provider access

from openai import OpenAI

client = OpenAI(base_url="http://127.0.0.1:8646/v1", api_key="ficelle-local")

# Ficelle selects among configured, enabled, eligible models.
# If none can serve the request, it returns an error.
client.chat.completions.create(
    model="ficelle/auto-tools",
    messages=[{"role": "user", "content": "..."}],
)

See what Ficelle Pro includes →

Pick a model type

Choose a model type for your request. Eligibility depends on the installed provider configuration, catalog data, free-access evidence and capability checks. A catalog listing alone does not confirm that your account can call a model.

• Tool calling
• JSON & structured output
• Fast, low-latency
• Reasoning
• Coding assistants
• Agent orchestration
• Long context
• Context compression
• Vision
• Audio input
• Video input
• Multimodal

Questions