{"format":"sfai.catalog-model.v1","dataQuality":{"version":"sfaie.catalog-quality.v1","status":"ok","quarantinedOffers":0,"quarantinedPerformanceOffers":0,"duplicateObservations":0,"issueCount":0,"issues":[],"issuesTruncated":false},"coverage":{"models":15226,"offers":16988,"providers":296,"priced":11103,"endpointModels":381,"independentlyBenchmarked":0},"units":{"version":"sfaie.catalog-units.v1","currency":"Offer.currency (ISO 4217)","tokenPrices":"currency per 1,000,000 tokens (input, output, cacheRead, cacheWrite when unit is \"1M tokens\")","unitPrice":"currency per Offer.unit","latency":"s (legacy field for older clients; prefer latencyMs)","latencyMs":"ms","latencyMetric":"p50 time to first token over Offer.performanceWindow","throughput":"tokens/s","throughputMetric":"p50 output tokens per second over Offer.performanceWindow","uptime":"percent","context":"tokens"},"policy":{"paidInference":false,"paidFeeds":false,"refreshMinutes":5,"historyDays":30,"benchmarkStatus":"Paid measurements are disabled. Published benchmarks and performance retain source attribution."},"canonical":true,"slug":"nvidia-nemotron-3-ultra-550b-a55b-nrctbx","sources":[{"id":"openrouter","count":470,"label":"OpenRouter","status":"ok","message":null,"observedAt":"2026-10-17T14:13:14.864Z","attemptedAt":"2026-10-17T14:13:14.356Z","quarantinedOffers":0}],"model":{"id":"nvidia/nemotron-3-ultra-550b-a55b","name":"NVIDIA: Nemotron 3 Ultra","aliases":["NVIDIA: Nemotron 3 Ultra"],"context":262144,"medusaId":"nvidia/nemotron-3-ultra-550b-a55b","released":"2026-06-04T05:33:28.000Z","developer":"nvidia","parameters":["frequency_penalty","include_reasoning","logit_bias","max_tokens","min_p","presence_penalty","reasoning","reasoning_effort","repetition_penalty","response_format","seed","stop","structured_outputs","temperature","tool_choice","tools","top_k","top_p"],"description":"NVIDIA Nemotron 3 Ultra is an open frontier-reasoning and orchestration model from NVIDIA, with 55B active parameters out of 550B total (MoE). Built on a hybrid Transformer-Mamba mixture-of-experts architecture, it...","benchmarkKey":"nvidia/nemotron-3-ultra-550b-a55b-20260604","sourceIdentity":{"source":"openrouter","version":"sfaie.source-model-identity.v1","sourceId":"nvidia/nemotron-3-ultra-550b-a55b","reference":"https://openrouter.ai/api/v1/models","observedAt":"2026-10-17T14:13:14.864Z","canonicalId":"nvidia/nemotron-3-ultra-550b-a55b-20260604","expirationDate":null},"inputModalities":["text"],"outputModalities":["text"],"offers":[{"id":"openrouter:nvidia/nemotron-3-ultra-550b-a55b:baseten/fp4","tier":"standard","unit":"1M tokens","input":0.6,"output":2.4,"region":null,"source":"openrouter","status":"priced","uptime":100,"channel":"openrouter","context":202800,"latency":0.338,"modelId":"nvidia/nemotron-3-ultra-550b-a55b","currency":"USD","provider":"BaseTen","cacheRead":0.12,"maxOutput":182520,"retention":null,"sourceUrl":"https://openrouter.ai/nvidia/nemotron-3-ultra-550b-a55b","unitPrice":null,"cacheWrite":null,"observedAt":"2026-10-17T13:32:26.666Z","parameters":["reasoning","include_reasoning","max_tokens","temperature","top_p","stop","frequency_penalty","presence_penalty","top_k","tools","tool_choice","reasoning_effort"],"provenance":"aggregator","throughput":100.5,"pricingRules":null,"quantization":"fp4","performanceWindow":"Previous 30 minutes; reported by OpenRouter","latencyMs":338},{"id":"openrouter:nvidia/nemotron-3-ultra-550b-a55b:baseten/fp4/us","tier":"standard","unit":"1M tokens","input":0.6,"output":2.4,"region":"us","source":"openrouter","status":"priced","uptime":100,"channel":"openrouter","context":202800,"latency":0.424,"modelId":"nvidia/nemotron-3-ultra-550b-a55b","currency":"USD","provider":"BaseTen","cacheRead":0.12,"maxOutput":182520,"retention":null,"sourceUrl":"https://openrouter.ai/nvidia/nemotron-3-ultra-550b-a55b","unitPrice":null,"cacheWrite":null,"observedAt":"2026-10-17T13:32:26.666Z","parameters":["reasoning","include_reasoning","max_tokens","temperature","top_p","stop","frequency_penalty","presence_penalty","top_k","tools","tool_choice","reasoning_effort"],"provenance":"aggregator","throughput":109,"pricingRules":null,"quantization":"fp4","performanceWindow":"Previous 30 minutes; reported by OpenRouter","latencyMs":424},{"id":"openrouter:nvidia/nemotron-3-ultra-550b-a55b:deepinfra/fp4","tier":"standard","unit":"1M tokens","input":0.5,"output":2.2,"region":null,"source":"openrouter","status":"priced","uptime":100,"channel":"openrouter","context":262144,"latency":1.8315,"modelId":"nvidia/nemotron-3-ultra-550b-a55b","currency":"USD","provider":"DeepInfra","cacheRead":0.1,"maxOutput":16384,"retention":null,"sourceUrl":"https://openrouter.ai/nvidia/nemotron-3-ultra-550b-a55b","unitPrice":null,"cacheWrite":null,"observedAt":"2026-10-17T13:32:26.666Z","parameters":["reasoning","include_reasoning","max_tokens","temperature","top_p","stop","frequency_penalty","presence_penalty","repetition_penalty","top_k","seed","min_p","tools","tool_choice","structured_outputs","response_format","logit_bias","reasoning_effort"],"provenance":"aggregator","throughput":35,"pricingRules":null,"quantization":"fp4","performanceWindow":"Previous 30 minutes; reported by OpenRouter","latencyMs":1831.5},{"id":"openrouter:nvidia/nemotron-3-ultra-550b-a55b:deepinfra/fp4/flex","tier":"flex","unit":"1M tokens","input":0.4,"output":1.76,"region":null,"source":"openrouter","status":"priced","uptime":null,"channel":"openrouter","context":262144,"latency":3.072,"modelId":"nvidia/nemotron-3-ultra-550b-a55b","currency":"USD","provider":"DeepInfra","cacheRead":0.08,"maxOutput":16384,"retention":null,"sourceUrl":"https://openrouter.ai/nvidia/nemotron-3-ultra-550b-a55b","unitPrice":null,"cacheWrite":null,"observedAt":"2026-10-17T13:32:26.666Z","parameters":["reasoning","include_reasoning","max_tokens","temperature","top_p","stop","frequency_penalty","presence_penalty","repetition_penalty","top_k","seed","min_p","logit_bias","response_format","structured_outputs","tools","tool_choice","reasoning_effort"],"provenance":"aggregator","throughput":34,"pricingRules":null,"quantization":"fp4","performanceWindow":"Previous 30 minutes; reported by OpenRouter","latencyMs":3072},{"id":"openrouter:nvidia/nemotron-3-ultra-550b-a55b:venice/fp8","tier":"standard","unit":"1M tokens","input":0.625,"output":3.125,"region":null,"source":"openrouter","status":"priced","uptime":100,"channel":"openrouter","context":256000,"latency":2.092,"modelId":"nvidia/nemotron-3-ultra-550b-a55b","currency":"USD","provider":"Venice","cacheRead":0.1875,"maxOutput":32768,"retention":null,"sourceUrl":"https://openrouter.ai/nvidia/nemotron-3-ultra-550b-a55b","unitPrice":null,"cacheWrite":null,"observedAt":"2026-10-17T13:32:26.666Z","parameters":["reasoning","include_reasoning","max_tokens","temperature","top_p","stop","frequency_penalty","presence_penalty","top_k","response_format","tools","tool_choice","reasoning_effort"],"provenance":"aggregator","throughput":34,"pricingRules":null,"quantization":"fp8","performanceWindow":"Previous 30 minutes; reported by OpenRouter","latencyMs":2092}],"firstSeen":"2026-09-16T23:50:33.173Z","ticker":"NEMOTR3UP","headlineOfferId":"openrouter:nvidia/nemotron-3-ultra-550b-a55b:deepinfra/fp4","displayName":"Nemotron 3 Ultra","brandedName":"NVIDIA Nemotron 3 Ultra"}}