Nemotron 3 Ultra
A 550B parameter (55B active) open reasoning model from NVIDIA, built for long-running agent workflows.
text → text · made by NVIDIA
Free
Offered at no charge, which is not the same as cheap: a free lane carries a rate limit and can be withdrawn. The prices above are what you pay when it is not available to you.
- OpenRouter
Sold by 35 ways
| Seller | Lane | Rate |
|---|---|---|
| DigitalOcean cloud | standard | $0.9per Mtok in$1.7per Mtok out$0.18per Mtok cachedmodels.dev · read 2026-11-01 |
| Thinking Machines Lab api | training | $2.49per Mtok in$6.22per Mtok out$0.5per Mtok cachedtinker-docs.thinkingmachines.ai · read 2026-08-25 |
| Thinking Machines Lab api | training-256k | $3.32per Mtok in$8.3per Mtok out$0.66per Mtok cachedtinker-docs.thinkingmachines.ai · read 2026-08-25 |
| OpenRouter 7 ways | $0.6per Mtok in$2.4per Mtok out$0.12per Mtok cached | |
| OpenRouter aggregator | free | $0per Mtok in$0per Mtok outopenrouter.ai · read 2026-11-01 |
| OpenRouter aggregator | deepinfra/fp4/flex | $0.4per Mtok in$1.76per Mtok out$0.08per Mtok cachedopenrouter.ai · read 2026-10-16 |
| OpenRouter aggregator | deepinfra/fp4 | $0.5per Mtok in$2.2per Mtok out$0.1per Mtok cachedopenrouter.ai · read 2026-08-25 |
| OpenRouter aggregator | standard | $0.6per Mtok in$2.4per Mtok out$0.12per Mtok cachedopenrouter.ai · read 2026-10-12 |
| OpenRouter aggregator | baseten/fp4 | $0.6per Mtok in$2.4per Mtok out$0.12per Mtok cachedopenrouter.ai · read 2026-08-25 |
| OpenRouter aggregator | baseten/fp4/us | $0.6per Mtok in$2.4per Mtok out$0.12per Mtok cachedopenrouter.ai · read 2026-10-15 |
| OpenRouter aggregatornot on the seller's list today | batch | $0.6per Mtok in$3.6per Mtok out$0.2per Mtok cachedopenrouter.ai · read 2026-08-24 |
| OpenRouter aggregatornot on the seller's list today | together | $0.6per Mtok in$3.6per Mtok out$0.2per Mtok cachedopenrouter.ai · read 2026-08-25 |
| OpenRouter aggregator | venice/fp8 | $0.62per Mtok in$3.12per Mtok out$0.19per Mtok cachedopenrouter.ai · read 2026-08-25 |
| Requesty 3 ways | $0.33per Mtok in$1.68per Mtok out | |
| Requesty aggregatornot on the seller's list today | doubleword flex | $0.33per Mtok in$1.68per Mtok outrouter.requesty.ai · read 2026-08-25 |
| Requesty aggregator | doubleword | $0.37per Mtok in$1.87per Mtok outrouter.requesty.ai · read 2026-10-10 |
| Requesty aggregator | fireworks | $0.6per Mtok in$2.4per Mtok out$0.12per Mtok cachedrouter.requesty.ai · read 2026-08-28 |
| Requesty aggregator | nebius | $1per Mtok in$3per Mtok out$1per Mtok cachedrouter.requesty.ai · read 2026-08-28 |
| Nous Research aggregatornot on the seller's list today | batch | $0.48per Mtok in$2.88per Mtok out$0.16per Mtok cachedinference-api.nousresearch.com · read 2026-08-25 |
| Nous Research aggregatornot on the seller's list today | batch resold | $0.48per Mtok in$2.88per Mtok out$0.16per Mtok cachedinference-api.nousresearch.com · read 2026-08-25 |
| Nous Research aggregatornot on the seller's list today | resold | $0.48per Mtok in$2.88per Mtok out$0.16per Mtok cachedinference-api.nousresearch.com · read 2026-08-25 |
| Nous Research aggregator | standard | $0.6per Mtok in$2.4per Mtok out$0.12per Mtok cachedinference-api.nousresearch.com · read 2026-10-29 |
| DeepInfra aggregator | standard | $0.5per Mtok in$2.2per Mtok out$0.1per Mtok cachedapi.deepinfra.com · read 2026-10-31 |
| DeepInfra aggregatornot on the seller's list today | alt | $0.5per Mtok in$2.2per Mtok outapi.deepinfra.com · read 2026-08-25 |
| DeepInfra aggregatornot on the seller's list today | bf16 | $1per Mtok in$5per Mtok outapi.deepinfra.com · read 2026-08-25 |
| ElectronHub aggregator | standard | $0.5per Mtok in$2.5per Mtok outapi.electronhub.ai · read 2026-11-01 |
| Vercel AI Gateway aggregator | standard | $0.5per Mtok in$2.4per Mtok out$0.12per Mtok cachedvercel.com · read 2026-08-25 |
| Vercel AI Gateway aggregatornot on the seller's list today | sandbox | $0.5per Mtok in$2.4per Mtok outvercel.com · read 2026-08-25 |
| Baseten 0 ways | $0.6per Mtok in$2.4per Mtok out$0.12per Mtok cached | |
| Baseten aggregatornot on the seller's list today | standard | $0.6per Mtok in$2.4per Mtok out$0.12per Mtok cachedraw.githubusercontent.com · read 2026-11-01 |
| Fireworks AI 2 ways | $0.6per Mtok in$2.4per Mtok out$0.12per Mtok cached | |
| Fireworks AI aggregator | standard | $0.6per Mtok in$2.4per Mtok out$0.12per Mtok cachedmodels.dev · read 2026-11-01 |
| Fireworks AI aggregator | priority | $0.75per Mtok in$3per Mtok out$0.15per Mtok cachedraw.githubusercontent.com · read 2026-11-01 |
| Hugging Face Inference Providers aggregatornot on the seller's list today | fireworks-ai | $0.6per Mtok in$2.4per Mtok outrouter.huggingface.co · read 2026-08-25 |
| Hugging Face Inference Providers aggregatornot on the seller's list today | together | $0.6per Mtok in$3.6per Mtok outrouter.huggingface.co · read 2026-08-25 |
| Hugging Face Inference Providers aggregator | deepinfra | $1per Mtok in$5per Mtok outrouter.huggingface.co · read 2026-08-25 |
| Together AI aggregator | standard | $0.6per Mtok in$3.6per Mtok out$0.2per Mtok cachedmodels.dev · read 2026-11-01 |
| Venice AI aggregator | standard | $0.62per Mtok in$3.12per Mtok out$0.19per Mtok cachedmodels.dev · read 2026-11-01 |
| Nebius Token Factory aggregator | standard | $1per Mtok in$3per Mtok outtokenfactory.nebius.com · read 2026-08-25 |
Measured 23 standings
About
Nemotron 3 Ultra — a text model from NVIDIA, sold by 13 companies from $0.5 in and $2.2 out per million tokens, placed 94th of 326 on GPQA Diamond.
It takes text and returns text, with a context window of 1,000,000 tokens. It was published in June 2026, trained on material up to February 2026. Its sellers say it can reason step by step and call a tool. The catalogue files it under reasoning and agents. Thirteen companies sell it. The cheapest is $0.5 in and $2.2 out per million tokens at DeepInfra. Beside the standard rate there are faster, separately routed, regional and flex lanes. It has been measured on 23 boards, and stands best at 94th of 326 on GPQA Diamond.
Every current figure
- Maker
- NVIDIA
- Register
- model
- Takes
- text
- Returns
- text
- Context
- 1,000,000 tokens
- Longest answer
- 512,300 tokens
- Published
- June 2026
- Knowledge to
- February 2026
- Parameters
- 303 billion
- Licence
- not read
- Sellers
- 13
- Maker's own price
- not read
- Price
- $0.5 in and $2.2 out per million tokens — DeepInfra
- Boards
- 23
- Best place
- 94th of 326 — GPQA Diamond
Known as 10 names
nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55Bnvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16fireworks/nemotron-3-ultra-nvfp4nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B-NVFP4nvidia/Nemotron-3-Ultra-550b-a55bnvidia/nemotron-3-ultra-550b-a55b:batchdoubleword/nvidia-nemotron-3-ultradoubleword/nvidia-nemotron-3-ultra:flexnebius/nvidia/nemotron-3-ultra-550b-a55bnvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16:peft:262144