Approved catalogue_
Choose the model. Keep the contract.
Provider pricing stays visible. Routing and governance stay behind one API.
TypeSafeJev 1.13.0
jev-1.13.0Jev 1.13.0 provides typed AI judgments through PipeLLM. Use model ID jev-1.13.0 with /v1/systemone.
OpenAIGPT 6 Sol
gpt-6-solDeveloped by OpenAI as part of its GPT-6 family, GPT-6 Sol is positioned as a cost-effective, high-tier language model. It occupies an intermediate place in the lineup, designed to sit directly beneath the flagship GPT-6 Astra tier while providing more advanced capability than the rapid-performance GPT-6 Luna variant.Engineered for demanding professional applications, GPT-6 Sol targets workloads that require strong performance and precision alongside balanced operating expenses. Its positioning makes it an effective choice for complex business tasks that benefit from near-flagship intelligence with greater resource efficiency.
OpenAIGPT-6 Astra
gpt-6-astraGPT-6 Astra provides text generation through PipeLLM. Use model ID gpt-6-astra with /v1/chat/completions.
GoogleGemini 3.8 Flash
gemini-3.8-flashGemini 3.8 Flash extends Google's Flash family with a focus on software engineering, agent tasks, and complex reasoning. It accepts text, visual material, audio, and documents, bringing these inputs together in text-based responses.The model is suited to coding assistants, multimodal analysis, and workflows that require repeated reasoning and tool use. Its long-context support helps applications work with substantial documents and extended conversations.
AnthropicClaude Opus 5 5
claude-opus-5-5Claude Opus 5.5 is an artificial intelligence model developed by Anthropic, made available to developers through the Claude Platform. It forms part of the publisher's expanding suite of models designed to be deployed and integrated directly into software applications.Within the Claude ecosystem, the model aligns with products such as Claude Code and Claude Code Enterprise. This positioning provides developers with access to Anthropic's capabilities across enterprise workflows and code-focused development environments.
Z.AIGLM 5.3
glm-5.3GLM-5.3 is Z.ai's reasoning model for demanding software engineering and extended agent workflows. It processes text and is designed to work through complex problems while retaining substantial contextual information.The model suits code analysis, implementation tasks, and multi-step technical work. Its focus on reasoning and token efficiency makes it a candidate for applications that need sustained problem solving rather than brief, isolated responses.
AnthropicClaude Opus 5
claude-opus-5Claude Opus 5 is Anthropic’s flagship model for demanding reasoning, coding, and long-horizon agentic work. It is particularly strong at end-to-end software tasks, code review and bug finding, visual analysis...
Kimi AIKimi K3
kimi-k3Kimi K3 is a 2.8T parameter open-weight multimodal reasoning model from Moonshot AI. It is suited for complex coding, knowledge work, and long-horizon agentic workflows, and is particularly strong at...
DISCOUNT30% OFF
ByteDanceDeepseek V4 1 Flash
deepseek-v4-1-flashApproved for governed routing through Relay.
DISCOUNT30% OFF
DeepSeekDeepseek V4 Pro
deepseek-v4-proDeepSeek V4 Pro is a large-scale Mixture-of-Experts model aimed at advanced reasoning, coding, long-context analysis, and long-horizon agent workflows. It is suited for full-codebase analysis, multi-step automation, and large-scale information synthesis where capability and efficiency both matter.
OpenAIGPT 6.1 Sol
gpt-6.1-solApproved for governed routing through Relay.
OpenAIGPT 6 Luna
gpt-6-lunaApproved for governed routing through Relay.
ByteDanceSeedream 5.0 Lite
doubao-seedream-5-0-260128Seedream 5.0 Lite provides image generation through PipeLLM. Use model ID doubao-seedream-5-0-260128 with /v1/images/generations.
ByteDanceSeedance 2.0 Fast
doubao-seedance-2-0-fast-260128Seedance 2.0 Fast provides video generation through PipeLLM. Use model ID doubao-seedance-2-0-fast-260128 with /v2/videos.
ByteDanceSeedance 2.5
doubao-seedance-2-5-260628Seedance 2.5 provides video generation through PipeLLM. Use model ID doubao-seedance-2-5-260628 with /v2/videos.
ByteDanceSeedance 2.0 Mini
doubao-seedance-2-0-mini-260615Seedance 2.0 Mini provides video generation through PipeLLM. Use model ID doubao-seedance-2-0-mini-260615 with /v2/videos.
OpenAIGPT-5.4 Nano
gpt-5.4-nanoGPT-5.4 Nano provides text generation through PipeLLM. Use model ID gpt-5.4-nano with /v1/chat/completions.
GoogleGemini 3.5 Flash Lite
gemini-3.5-flash-liteGemini 3.5 Flash Lite provides text generation through PipeLLM. Use model ID gemini-3.5-flash-lite with /v1/chat/completions.
GoogleGemini 3.1 Flash Image
gemini-3.1-flash-imageGemini 3.1 Flash Image supports image generation and text through PipeLLM. Use model ID gemini-3.1-flash-image with the configured compatible endpoint.
CohereRerank v4.0 Pro
rerank-v4.0-proRerank v4.0 Pro provides document reranking through PipeLLM. Use model ID rerank-v4.0-pro with /v2/rerank.
CohereRerank v4.0 Fast
rerank-v4.0-fastRerank v4.0 Fast provides document reranking through PipeLLM. Use model ID rerank-v4.0-fast with /v2/rerank.
ByteDanceSeedance 2.0
seedance-2.0Seedance 2.0 generates video through PipeLLM. Use model seedance-2.0 with POST /v2/videos.
ByteDanceDoubao Seed 2.0 Mini
doubao-seed-2-0-mini-260428Doubao Seed 2.0 Mini provides text generation through PipeLLM. Use model ID doubao-seed-2-0-mini-260428 with /v3/chat/completions.
OpenAIGPT-5.2
gpt-5.2GPT-5.2 provides text generation through PipeLLM. Use model ID gpt-5.2 with /v1/chat/completions.