InferenceChoose a Provider and Model

Choose a Model

View as Markdown

Use the curated model choices as starter guidance when selecting a cloud model during onboarding. The provider catalog remains authoritative for exact context-window limits and current pricing. Runtime route validation determines current availability because catalog entries can outlive their backing endpoints.

Catalog Selection

During interactive NVIDIA Endpoints onboarding, NemoClaw loads NVIDIA’s public featured model catalog once per onboarding session and reports progress before displaying the model picker. OpenRouter uses the same catalog-backed picker flow and featured model list with its own provider route and credential check. NVIDIA Endpoints excludes NVIDIA-retired or unsafe choices and corrects known catalog lag before displaying the result.

If the catalog is unavailable, malformed, or contains no safe model IDs, the wizard warns you and uses its bundled fallback list. Nemotron 3 Super remains the shared default for OpenClaw and Hermes when it is present. LangChain Deep Agents Code uses Nemotron 3 Ultra as its NVIDIA Endpoints default. If an agent’s default is unavailable, the first live featured model becomes the interactive default.

If NEMOCLAW_MODEL contains a safe custom model ID that is absent from the live catalog, it does not replace the live menu default. Choose Other to use that value as the pre-filled manual entry. NemoClaw validates the manual entry against the selected provider before continuing. NemoClaw does not display or accept an unsafe value as the manual-entry prefill.

Model Task Fit

The relative labels compare models within the curated onboarding choices rather than across every model that a provider offers.

ModelBest forRelative latencyTool useContext fitRelative cost
nvidia/nemotron-3-ultra-550b-a55bQuality-sensitive reasoning, careful synthesis, and complex reviewsHigherStrong for complex tool plansLarge agent contextHigher
nvidia/nemotron-3-super-120b-a12bHosted agent work, multi-step planning, and tool-heavy shell workflowsMediumStrong default for OpenClaw tool loopsLarge agent contextMedium
minimaxai/minimax-m3Long-form writing, multi-turn assistant work, and broad instruction followingMediumGood for structured assistant turnsLarge agent contextMedium
gpt-5.4General OpenAI-backed agent work and high-quality reasoningMediumStrongLarge agent contextMedium to high
gpt-5.4-miniLatency-sensitive routine automation and repeated helper callsLowGoodMedium to large contextLow
gpt-5.4-nanoClassification, routing, extraction, and small helper tasksVery lowBasic to good for simple tool loopsMedium contextVery low
gpt-5.4-pro-2026-03-05Quality-first complex reasoning where latency and cost are secondaryHighestValidate Responses API support before long tool loopsLarge agent contextHighest
claude-sonnet-4-6Balanced coding, writing, analysis, and multi-step tool workMediumStrongLarge agent contextMedium to high
claude-haiku-4-5Fast summarization, routing, extraction, and lightweight assistant turnsLowGood for simple tool loopsMedium to large contextLow
claude-opus-4-6Deep analysis, careful writing, and quality-first planningHigherStrongLarge agent contextHigher
gemini-3.1-pro-previewLarge-context analysis, synthesis, and preview-feature evaluationMedium to highGood, with tool continuation validation for the selected routeExtensive contextMedium to high
gemini-3.1-flash-lite-previewLow-cost extraction, classification, and simple helper callsLowBasic to good for simple tool loopsMedium to large contextLow
gemini-3-flash-previewFast general assistant tasks and preview-feature evaluationLowGood for simple tool loopsLarge contextLow
gemini-2.5-proLarge-context analysis, long-document synthesis, and complex reasoningMedium to highGoodExtensive contextMedium to high
gemini-2.5-flashLatency-sensitive general assistant and multimodal tasksLowGood for simple tool loopsLarge contextLow
gemini-2.5-flash-liteLowest-cost helper calls, extraction, and classificationVery lowBasic to good for simple tool loopsMedium to large contextVery low

Nemotron Deployment Choice

Nemotron models expose OpenAI-compatible APIs across the supported deployment surfaces. Choose the onboarding option that matches the host.

Nemotron hostOnboarding option
NVIDIA-hosted on build.nvidia.comNVIDIA Endpoints
Self-hosted NIM containerOther OpenAI-compatible endpoint
Enterprise NVIDIA AI Enterprise gatewayOther OpenAI-compatible endpoint
vLLM, SGLang, or TRT-LLM serving Nemotron weightsOther OpenAI-compatible endpoint
Local NIM started by the wizardLocal NVIDIA NIM