NVIDIALlama 3.1 Nemotron 70B Instruct
Best configuration ranks #453 of 610 on the BenchLeader Index at 42.8 ±4.1. Last measured 2 Sept 2026. Released 15 Apr 2025.
- Blended price
- $1.20/M
- $1.20 in · $1.20 out
- Output speed
- 43 tok/s
- First answer
- 8.27 s
- first token 8.27 s
- Context
- 128k
- Overall index43
- Reasoning46
- Coding45
- Agents & tools38
- Knowledge45
- Instruction following34
- Human preference46
- Long context29
- Composite38
Benchmark results
One column per reasoning effort. Rank is among every configuration of every model on that benchmark. Hover a score for the run it came from.
Reasoning
| Benchmark | default | Source |
|---|---|---|
| LMArena Hard Prompts | 1310#231 | LMArena |
| GPQA Diamond (AA)not in index | 46.5%#423 | Artificial Analysis |
| Humanity's Last Exam (AA)not in index | 4.2%#446 | Artificial Analysis |
Coding
| Benchmark | default | Source |
|---|---|---|
| LMArena Coding | 1328#237 | LMArena |
Agents & tools
| Benchmark | default | Source |
|---|---|---|
| Terminal-Bench Hard | 4.5%#270 | Artificial Analysis |
| τ²-Bench Telecom (AA)not in index | 23.1%#294 | Artificial Analysis |
Knowledge
| Benchmark | default | Source |
|---|---|---|
| AA-Omniscience | -40.8#257 | Artificial Analysis |
Instruction following
| Benchmark | default | Source |
|---|---|---|
| IFBench | 30.8%#342 | Artificial Analysis |
Human preference
| Benchmark | default | Source |
|---|---|---|
| LMArena Text | 1299#233 | LMArena |
Long context
| Benchmark | default | Source |
|---|---|---|
| AA-LCR | 8.3%#400 | Artificial Analysis |
Composite
| Benchmark | default | Source |
|---|---|---|
| AA Intelligence Index | 6.9#420 | Artificial Analysis |
What a task costs
Estimates from list price, output speed and time to first answer for the best configuration. “With caching” assumes three-quarters of the input is served from the prompt cache. Reasoning tokens are not modelled.
| Workload | Tokens in / out | Cost | With caching | Time |
|---|---|---|---|---|
| Chat reply | 400 / 300 | $0.0008 | – | 15.3 s |
| Summarise a 30-page report | 12,000 / 600 | $0.015 | – | 22.3 s |
| Code edit | 6,000 / 1,500 | $0.0090 | – | 43.3 s |
| Agentic coding session | 60,000 / 4,000 | $0.077 | – | 1.7 min |
| Structured extraction | 2,000 / 200 | $0.0026 | – | 12.9 s |
See also
Data as of 9 Sept 2026. Compare with another model.