Largest Nemotron 3 model for maximum open-weight reasoning and agent accuracyNemotron 3 Ultra 550B A55B by nvidia is offered by 34 providers on this page. Public prices are shown for 24 of them. Its official list price is $0 per 1M tokens. No paid channel is priced below the official list price. Prices vary about 18× across the providers that host it. It also has 10 free ($0) channels; free tiers usually carry rate limits, and subscription-covered access bills $0 per token only after the subscription fee.
Artificial Analysis rates it 11.8 on the Intelligence Index, with 24.2 for Finance & Accounting, its strongest domain. Against its lowest blended price of $0.10 per 1M, that is roughly 118 index points per dollar, which is the value ratio the leaderboards rank on. Median output speed is 255 tokens per second, with 1.03s to the first token. Running one task of the Artificial Analysis suite costs about $2.1897, which reflects how many tokens its reasoning consumes rather than the unit price alone.
The context window is 1,000,000 tokens at the reference host, but hosts report different limits, from 65,536 to 1,048,576, so the usable window depends on the provider you pick. It supports reasoning and tool use. The weights are open, so it can also be self-hosted or served through a gateway of your choice. Providers report the capability flags inconsistently, so verify a specific feature against the host you plan to use. Its training knowledge cuts off at 2025-09.
Quality is independently evaluated by Artificial Analysis. Speed/latency are model-level medians.
| Provider | Tier | Input | Output | Cache read | Cache write | Context | Output limit | Status |
|---|---|---|---|---|---|---|---|---|
| NvidiaOfficial | Cloud | Free | — | — | 1,000,000 | 65,536 | ||
| Requesty | Gateway | Free | — | — | 1,048,576 ⚠ | 65,536 | ||
| OpenRouter nvidia/nemotron-3-ultra-550b-a55b:free | Gateway | Free | — | — | 1,000,000 | 65,536 | ||
| Bothub nemotron-3-ultra-550b-a55b:free | Gateway | Free | — | — | 1,000,000 | 128,000 | ||
| AIHubMix nemotron-3-ultra-550b-a55b-free | Gateway | Free | — | — | 1,000,000 | 128,000 | ||
| Kilo Gateway nvidia/nemotron-3-ultra-550b-a55b:free | Gateway | Free | — | — | 1,000,000 | 65,536 | ||
| Kenari nemotron-3-ultra-550b-a55b:free | Gateway | Free | — | — | 1,000,000 | 128,000 | ||
| InferAll | Gateway | Free | — | — | 1,000,000 | 128,000 | ||
| UnoRouter nemotron-3-ultra-550b-a55b:free | Gateway | Free | — | — | 1,000,000 | 128,000 | ||
| OpenCode nemotron-3-ultra-free | Gateway | Free | $0 | — | 1,000,000 | 128,000 | deprecated |
Sorted by blended price (input×0.75 + output×0.25) asc. The official channel always shows regardless of rank. Whether a gateway's low price is actually usable can't be verified.
Interleaved thinking (reasoning between tool calls) is declared by 6 of 34 providers.
7 / 34 providers expose no reasoning control (reasoning_options: []).
10 more channels offer $0 (Nvidia, Requesty, OpenRouter etc.); free tiers usually have rate limits and no SLA, excluded from ranking.
| Name | Conditions | Score | Metric | Source |
|---|---|---|---|---|
| SWE-Bench Verified | — | 70.7 | resolved | Source ↗ |
| SWE-Bench Multilingual | — | 67.7 | resolve rate | Source ↗ |
| Terminal-Bench | v2.1 | 56.4 | success rate | Source ↗ |
| GPQA | variant: no tools | 87 | accuracy | Source ↗ |
| Humanity's Last Exam | variant: no tools | 26.7 | accuracy | Source ↗ |
| Humanity's Last Exam | variant: with tools | 37.4 | accuracy | Source ↗ |
| LiveCodeBench | vv6 | 89 | pass@1 | Source ↗ |
| MMLU-Pro | — | 86.8 | accuracy | Source ↗ |
| BrowseComp | — | 44.4 | accuracy | Source ↗ |
| IFBench | variant: prompt loose | 81.7 | accuracy | Source ↗ |
| GDPval | — | 46.7 | wins or ties | Source ↗ |
The same benchmark scores very differently across harness / dataset, so the qualifying conditions must be shown together.
Individual evaluations run by Artificial Analysis, on the same reasoning tier as the intelligence score above. Each benchmark has its own task set and scale (percent correct, points or an index), so rows are not comparable with one another.