NVIDIALanguage model
Mistral Nemotron
Mistral Nemotron has no overall index value yet because there are too few independent benchmark results.
3out of 100same as last month
based on two data groups, mostly search
Summary
Mistral Nemotron launched in June 2025 as a joint NVIDIA-Mistral model for enterprise agents. It is a specific enterprise checkpoint, not a name for the Mistral family as a whole.
Access runs through a downloadable NVIDIA NIM microservice. NVIDIA documents both on-premises and cloud deployment. The NIM serving package is therefore part of the documented model form.
All benchmark results
Every published result with its source, test setup and comparison models.
Show all 14 published benchmark results
Page 1 of 2
Measurements without matching peer values
These measurements have no matching peer values under the same test conditions. Their original values and sources remain available here.
HumanEval (Show measurement, test conditions, and source)
- Source value
- 92.68
- Score
- 92.68
- Metric
- Instruct, 0-shot, pass@1
- Unit
- No unit provided
- Benchmark version
- Mistral-Nemotron NIM model version v1
- Category
- coding
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- NVIDIA NIM Mistral-Nemotron model card v1
- Status
- active
- Retrieved at
- 2026-10-17
- Methodology
- Vendor-reported result. NVIDIA NIM card v1, Coding & Programming table. Instruct, 0-shot, pass@1. The card lists H100 as test hardware and TensorRT-LLM and vLLM as engines but does not map a specific engine to this row. Unit, task N, confidence interval, and reasoning mode are not reported.
- Context
- The card labels values only as Score and supplies no unit. NVIDIA's NIM listing captured on 2026-10-17 says the endpoint is no longer supported after 2026-09-28. These are historical vendor-reported results, not a current endpoint measurement. The card gives version v1 but no immutable weight digest.
LiveCodeBench v6 (Show measurement, test conditions, and source)
- Source value
- 27.42
- Score
- 27.42
- Metric
- 0-shot; the card labels the value only as Score and does not name a more specific metric
- Unit
- No unit provided
- Benchmark version
- LiveCodeBench v6; Mistral-Nemotron NIM model version v1
- Category
- coding
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- NVIDIA NIM Mistral-Nemotron model card v1
- Status
- active
- Retrieved at
- 2026-10-17
- Methodology
- Vendor-reported result. NVIDIA NIM card v1, Coding & Programming table. LiveCodeBench v6, 0-shot. The card lists H100 as test hardware and TensorRT-LLM and vLLM as engines but does not map a specific engine to this row. Unit, task N, confidence interval, and reasoning mode are not reported.
- Context
- The card labels values only as Score and supplies no unit. NVIDIA's NIM listing captured on 2026-10-17 says the endpoint is no longer supported after 2026-09-28. These are historical vendor-reported results, not a current endpoint measurement. The card gives version v1 but no immutable weight digest.
IfEval (Show measurement, test conditions, and source)
- Source value
- 87.33
- Score
- 87.33
- Metric
- 0-shot
- Unit
- No unit provided
- Benchmark version
- Mistral-Nemotron NIM model version v1
- Category
- instruction-following
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- NVIDIA NIM Mistral-Nemotron model card v1
- Status
- active
- Retrieved at
- 2026-10-17
- Methodology
- Vendor-reported result. NVIDIA NIM card v1, Instruction Following table. 0-shot. The card lists H100 as test hardware and TensorRT-LLM and vLLM as engines but does not map a specific engine to this row. Unit, task N, confidence interval, and reasoning mode are not reported.
- Context
- The card labels values only as Score and supplies no unit. NVIDIA's NIM listing captured on 2026-10-17 says the endpoint is no longer supported after 2026-09-28. These are historical vendor-reported results, not a current endpoint measurement. The card gives version v1 but no immutable weight digest.
MATH (Show measurement, test conditions, and source)
- Source value
- 91.14
- Score
- 91.14
- Metric
- Instruct, 0-shot
- Unit
- No unit provided
- Benchmark version
- Mistral-Nemotron NIM model version v1
- Category
- mathematics
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- NVIDIA NIM Mistral-Nemotron model card v1
- Status
- active
- Retrieved at
- 2026-10-17
- Methodology
- Vendor-reported result. NVIDIA NIM card v1, Mathematics table. Instruct, 0-shot. The card lists H100 as test hardware and TensorRT-LLM and vLLM as engines but does not map a specific engine to this row. Unit, task N, confidence interval, and reasoning mode are not reported.
- Context
- The card labels values only as Score and supplies no unit. NVIDIA's NIM listing captured on 2026-10-17 says the endpoint is no longer supported after 2026-09-28. These are historical vendor-reported results, not a current endpoint measurement. The card gives version v1 but no immutable weight digest.
MMLU Pro (Show measurement, test conditions, and source)
- Source value
- 73.81
- Score
- 73.81
- Metric
- Instruct, 5-shot, chain-of-thought (CoT)
- Unit
- No unit provided
- Benchmark version
- Mistral-Nemotron NIM model version v1
- Category
- knowledge-reasoning
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- NVIDIA NIM Mistral-Nemotron model card v1
- Status
- active
- Retrieved at
- 2026-10-17
- Methodology
- Vendor-reported result. NVIDIA NIM card v1, General Knowledge & Reasoning table. Instruct, 5-shot CoT. The card lists H100 as test hardware and TensorRT-LLM and vLLM as engines but does not map a specific engine to this row. Unit, task N, and confidence interval are not reported.
- Context
- The card labels values only as Score and supplies no unit. NVIDIA's NIM listing captured on 2026-10-17 says the endpoint is no longer supported after 2026-09-28. These are historical vendor-reported results, not a current endpoint measurement. The card gives version v1 but no immutable weight digest.
MMLU (English) (Show measurement, test conditions, and source)
- Source value
- 84.84
- Score
- 84.84
- Metric
- Instruct, 5-shot
- Unit
- No unit provided
- Benchmark version
- Mistral-Nemotron NIM model version v1
- Category
- multilingual
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- NVIDIA NIM Mistral-Nemotron model card v1
- Status
- active
- Retrieved at
- 2026-10-17
- Methodology
- Vendor-reported result. NVIDIA NIM card v1, MMLU by Language table, English row. Instruct, 5-shot. The card lists H100 as test hardware and TensorRT-LLM and vLLM as engines but does not map a specific engine to this row. Unit, task N, confidence interval, and reasoning mode are not reported.
- Context
- The card labels values only as Score and supplies no unit. NVIDIA's NIM listing captured on 2026-10-17 says the endpoint is no longer supported after 2026-09-28. These are historical vendor-reported results, not a current endpoint measurement. The card gives version v1 but no immutable weight digest.
CMMLU (Chinese) (Show measurement, test conditions, and source)
- Source value
- 80.54
- Score
- 80.54
- Metric
- Chinese, Instruct, 5-shot
- Unit
- No unit provided
- Benchmark version
- Mistral-Nemotron NIM model version v1
- Category
- multilingual
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- NVIDIA NIM Mistral-Nemotron model card v1
- Status
- active
- Retrieved at
- 2026-10-17
- Methodology
- Vendor-reported result. NVIDIA NIM card v1, MMLU by Language table, Chinese row. Instruct, 5-shot. The card lists H100 as test hardware and TensorRT-LLM and vLLM as engines but does not map a specific engine to this row. Unit, task N, confidence interval, and reasoning mode are not reported.
- Context
- The card labels values only as Score and supplies no unit. NVIDIA's NIM listing captured on 2026-10-17 says the endpoint is no longer supported after 2026-09-28. These are historical vendor-reported results, not a current endpoint measurement. The card gives version v1 but no immutable weight digest.
JMMLU (Japanese) (Show measurement, test conditions, and source)
- Source value
- 80.85
- Score
- 80.85
- Metric
- Japanese, Instruct, 5-shot
- Unit
- No unit provided
- Benchmark version
- Mistral-Nemotron NIM model version v1
- Category
- multilingual
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- NVIDIA NIM Mistral-Nemotron model card v1
- Status
- active
- Retrieved at
- 2026-10-17
- Methodology
- Vendor-reported result. NVIDIA NIM card v1, MMLU by Language table, Japanese row. Instruct, 5-shot. The card lists H100 as test hardware and TensorRT-LLM and vLLM as engines but does not map a specific engine to this row. Unit, task N, confidence interval, and reasoning mode are not reported.
- Context
- The card labels values only as Score and supplies no unit. NVIDIA's NIM listing captured on 2026-10-17 says the endpoint is no longer supported after 2026-09-28. These are historical vendor-reported results, not a current endpoint measurement. The card gives version v1 but no immutable weight digest.
KMMLU (Korean) (Show measurement, test conditions, and source)
- Source value
- 64.56
- Score
- 64.56
- Metric
- Korean, Instruct, 5-shot
- Unit
- No unit provided
- Benchmark version
- Mistral-Nemotron NIM model version v1
- Category
- multilingual
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- NVIDIA NIM Mistral-Nemotron model card v1
- Status
- active
- Retrieved at
- 2026-10-17
- Methodology
- Vendor-reported result. NVIDIA NIM card v1, MMLU by Language table, Korean row. Instruct, 5-shot. The card lists H100 as test hardware and TensorRT-LLM and vLLM as engines but does not map a specific engine to this row. Unit, task N, confidence interval, and reasoning mode are not reported.
- Context
- The card labels values only as Score and supplies no unit. NVIDIA's NIM listing captured on 2026-10-17 says the endpoint is no longer supported after 2026-09-28. These are historical vendor-reported results, not a current endpoint measurement. The card gives version v1 but no immutable weight digest.
MMLU (French) (Show measurement, test conditions, and source)
- Source value
- 82.99
- Score
- 82.99
- Metric
- French, 5-shot
- Unit
- No unit provided
- Benchmark version
- Mistral-Nemotron NIM model version v1
- Category
- multilingual
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- NVIDIA NIM Mistral-Nemotron model card v1
- Status
- active
- Retrieved at
- 2026-10-17
- Methodology
- Vendor-reported result. NVIDIA NIM card v1, MMLU by Language table, French row (Fr MMLU). 5-shot. The card lists H100 as test hardware and TensorRT-LLM and vLLM as engines but does not map a specific engine to this row. Unit, task N, confidence interval, and reasoning mode are not reported.
- Context
- The card labels values only as Score and supplies no unit. NVIDIA's NIM listing captured on 2026-10-17 says the endpoint is no longer supported after 2026-09-28. These are historical vendor-reported results, not a current endpoint measurement. The card gives version v1 but no immutable weight digest.
MMLU (German) (Show measurement, test conditions, and source)
- Source value
- 81.99
- Score
- 81.99
- Metric
- German, 5-shot
- Unit
- No unit provided
- Benchmark version
- Mistral-Nemotron NIM model version v1
- Category
- multilingual
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- NVIDIA NIM Mistral-Nemotron model card v1
- Status
- active
- Retrieved at
- 2026-10-17
- Methodology
- Vendor-reported result. NVIDIA NIM card v1, MMLU by Language table, German row (De MMLU). 5-shot. The card lists H100 as test hardware and TensorRT-LLM and vLLM as engines but does not map a specific engine to this row. Unit, task N, confidence interval, and reasoning mode are not reported.
- Context
- The card labels values only as Score and supplies no unit. NVIDIA's NIM listing captured on 2026-10-17 says the endpoint is no longer supported after 2026-09-28. These are historical vendor-reported results, not a current endpoint measurement. The card gives version v1 but no immutable weight digest.
MMLU (Spanish) (Show measurement, test conditions, and source)
- Source value
- 83.61
- Score
- 83.61
- Metric
- Spanish, 5-shot
- Unit
- No unit provided
- Benchmark version
- Mistral-Nemotron NIM model version v1
- Category
- multilingual
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- NVIDIA NIM Mistral-Nemotron model card v1
- Status
- active
- Retrieved at
- 2026-10-17
- Methodology
- Vendor-reported result. NVIDIA NIM card v1, MMLU by Language table, Spanish row (Es MMLU). 5-shot. The card lists H100 as test hardware and TensorRT-LLM and vLLM as engines but does not map a specific engine to this row. Unit, task N, confidence interval, and reasoning mode are not reported.
- Context
- The card labels values only as Score and supplies no unit. NVIDIA's NIM listing captured on 2026-10-17 says the endpoint is no longer supported after 2026-09-28. These are historical vendor-reported results, not a current endpoint measurement. The card gives version v1 but no immutable weight digest.
Page 1 of 2
Popularity compared
How often the model is searched for, used and mentioned, compared with every model of its category.
Popularity index
The 100 most popular models of this category, 0 to 100, higher is more popular
20 of 101 models
As of November 2026 · Source: gradually.ai popularity index
Over time
Weekly values from October 18, 2026 to November 7, 2026
Mistral NemotronMuse Wonder 3GPT-6.1 Terra 2Falcon-Emirati 0Gemini 4 Carbon 0
Method changes in the timeline, a jump at these points is no market event:
- 1November 1, 2026: Value only with search or usage data, 30-day attention
What is gaining popularity right now, across models, vendors, harnesses and AI subscriptions, is collected in All AI trends.
Similar models
More models of the same kind as Mistral Nemotron, chosen by model line, developer, size, price and release date.
- Llama 3.3 Nemotron Super 49B v1.5NVIDIA · same developer
- Nemotron Nano 2 9BNVIDIA · same developer
- Llama 3.1 Nemotron 70B InstructNVIDIA · same developer
- Llama 3.3 Nemotron Super 49B v1NVIDIA · same developer
- Llama 3.1 Nemotron Ultra 253BNVIDIA · same developer
- Nemotron 3 NanoNVIDIA · same developer
- Nemotron Nano 2 VL 12BNVIDIA · same developer
- Nemotron 3 SuperNVIDIA · same developer
All specifications
Input and output
Watermarking and provenance
AI watermarking refers to invisible signals embedded in the output. Visible marks and provenance metadata are listed separately. Unknown means there is insufficient evidence for that particular field.
- AI watermarking
- Unknown
- no source
- Watermarking technique
- Not known
- no source
Licence and access
Further details
- NIM endpoint notice
- The captured NIM page says the API will be deprecated on 2026-09-27 and will no longer be supported after 2026-09-28. (retrieved 2026-10-17)
- Source: NIM endpoint notice
Frequently asked questions about Mistral Nemotron
How large is the context window of Mistral Nemotron?
128,000 tokens. That is how much text Mistral Nemotron reads at once, about 190 A4 pages.
Page count as a rule of thumb from Artificial Analysis
Can I run Mistral Nemotron locally?
Yes, the weights are open.
Stated by NVIDIA
- Open weights
- yes
- License
- NVIDIA Open Model License
- Parameters
- n/a
- active per token
- n/a
Sources and data date
Show all 11 sources
Benchmark sources at a glance
Which independent evaluators tested this model. This is coverage, not a quality ranking.
No active result from the reviewed external evaluators can currently be attributed to this model. Other source evidence remains available below.
Without vendor claims and historical or disputed results; those stay labelled in the single values.
| Type | Evidence and data date |
|---|---|
| Input | Documented modalities. Retrieved October 27, 2026. |
| Output | Documented modalities. Retrieved October 27, 2026. |
| Research status | Research date October 15, 2026. 5 source URLs checked. This documents the inspected sources, not an exhaustive inventory of every publication. |
| Unresolved | exact-checkpoint-identity Note The NVIDIA NIM model card identifies the mistral-nemotron v1 endpoint and reports vendor benchmark scores for that endpoint. It does not publish an immutable checkpoint or weight digest, so the served v1 alias is not mapped to exact public model weights. Mistral-Nemotron-Vision and Mistral NeMo/Minitron remain distinct models and are not substitutes. |
| Unresolved | quality-benchmarks Note NVIDIA's NIM model card for mistral-nemotron version v1 reports 14 numerical benchmark scores, including HumanEval, LiveCodeBench v6, IfEval, MATH, MMLU Pro and language-specific MMLU rows. The card labels cells only as Score and does not publish per-row units, sample counts or confidence intervals. It lists H100, TensorRT-LLM, vLLM and reasoning support generally, without mapping engines or reasoning modes to rows except the MMLU Pro CoT label. No immutable checkpoint digest is published, and the NIM alias is discontinued. |
| Not published | score-unit-task-n-confidence-interval Original wording The model card labels the cells only as Score and publishes no unit, per-task denominator, sample count, or confidence interval. These are not inferred. |
| Not published | per-row-engine-and-reasoning-mode Original wording The card lists H100 hardware, TensorRT-LLM and vLLM, and general reasoning support, but does not map engine or reasoning mode to individual rows. Only the explicitly named MMLU Pro row says CoT. |
| Not published | immutable-checkpoint-digest Original wording The card gives model version v1 but no immutable checkpoint or weight digest. The NIM listing identifies the endpoint alias. |
| Additional source | NVIDIA NIM Mistral-Nemotron model card v1 (retrieved October 17, 2026) |
| Additional source | Model metadata (retrieved October 16, 2026; October 17, 2026) · NVIDIA Mistral Nemotron Build listing · Editorial description reviewed October 2, 2026 |
| Additional source | Model metadata (retrieved October 16, 2026) · NVIDIA Mistral Nemotron announcement · Editorial description reviewed October 2, 2026 |
| Additional source | NVIDIA Mistral Nemotron NIM model card · Editorial description reviewed October 2, 2026 |
| Additional source | OpenRouter: Popularity index (derived), CC BY 4.0 |
| Additional source | Vercel AI Gateway: Popularity index (derived), CC BY 4.0 |
| Additional source | Ollama: cumulative downloads for the popularity index |
| Additional source | Replicate: cumulative runs for the popularity index |
| Additional source | Google Trending Now: limited feed selection for Germany and the US |
| Additional source | Creative Commons Attribution 4.0 |