b-ai · Flagship
GPT-5.5
Highest quality general model for advanced reasoning and production workloads.
gpt-5.5
Model catalog
Choose models by use case, not upstream provider keys.
Customers see clean SYC API model names. Your upstream keys stay private in the server environment and LiteLLM routes requests behind the scenes.
b-ai · Fast flagship
GPT-5.5 Instant
Fast premium model for interactive apps, agents, and chat products.
gpt-5.5-instant
b-ai · Pro
GPT-5.4 Pro
Stronger reasoning tier for coding, analysis, and automation workflows.
gpt-5.4-pro
b-ai · Balanced
GPT-5.4
Balanced quality and latency for SaaS features and developer tools.
gpt-5.4
b-ai · Balanced
GPT-5.2
Reliable default model for chat, summarization, and app integrations.
gpt-5.2
b-ai · Efficient
GPT-5.4 Mini
Cost-efficient default for support bots, tools, and high-volume calls.
gpt-5.4-mini
b-ai · Efficient
GPT-5 Mini
Lightweight model for fast responses and budget-sensitive workloads.
gpt-5-mini
b-ai · Low cost
GPT-5.4 Nano
Very low-cost model for extraction, classification, and simple automations.
gpt-5.4-nano
b-ai · Low cost
GPT-5 Nano
Smallest GPT option for simple routing, tagging, and utility calls.
gpt-5-nano
b-ai · Premium reasoning
Claude Opus 4.7
Premium Claude-style reasoning for complex writing, code, and analysis.
claude-opus-4.7
b-ai · Premium reasoning
Claude Opus 4.5
Claude Opus tier for writing, planning, and detailed code review.
claude-opus-4.5
b-ai · Reasoning
Claude Sonnet 4.6
Balanced Claude option for coding, writing, and agent workflows.
claude-sonnet-4.6
b-ai · Reasoning
Claude Sonnet 4.5
General Claude-style reasoning model for product and developer use cases.
claude-sonnet-4.5
b-ai · Fast
Claude Haiku 4.5
Fast Claude option for lightweight tasks and high-throughput apps.
claude-haiku-4.5
b-ai · Long context
Gemini 3.1 Pro
Gemini-style model for long context, research, and multimodal workflows.
gemini-3.1-pro
b-ai · Fast
Gemini 3 Flash
Fast Gemini option for latency-sensitive applications.
gemini-3-flash
b-ai · Reasoning
DeepSeek V4 Pro
DeepSeek pro model for coding, reasoning, and structured output.
deepseek-v4-pro
b-ai · Fast
DeepSeek V4 Flash
Fast DeepSeek option for code assistance and automation.
deepseek-v4-flash
b-ai · Balanced
DeepSeek V3.2
Balanced DeepSeek model for chat, code, and general usage.
deepseek-v3.2
b-ai · General
GLM 5.1
GLM model for multilingual and general assistant workloads.
glm-5.1
b-ai · General
GLM 5
General GLM model for chat, tools, and automation.
glm-5
b-ai · Long context
Kimi K2.5
Kimi-style model for long context reading and document workflows.
kimi-k2.5
b-ai · General
MiniMax M2.7
MiniMax model for chat, content, and application workflows.
minimax-m2.7
b-ai · General
MiniMax M2.5
MiniMax model for general purpose AI product features.
minimax-m2.5
How upstream keys connect
Put your internal New API token in .env, then LiteLLM reads it from litellm/config.yaml. Customers only receive virtual SYC API keys.
NEWAPI_BASE_URL=https://sycapi.com/v1
NEWAPI_API_KEY=your-internal-newapi-token
LITELLM_MASTER_KEY=sk-syc-admin-change-me