Model comparison
Llama 3.1-70B vs Mistral Medium
Mistral Medium is the stronger model overall, scoring 36.0 to 29.4 on the Noometry Index. Llama 3.1-70B costs 7.5× less per token, which makes it the better buy when Mistral Medium's lead doesn't matter for your workload.
Last verified . 23 shared benchmarks.
Meta
Mistral AISummary
- They share 23 benchmarks with published results for both. Llama 3.1-70B scores higher in 0 categories and Mistral Medium in 9 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where Mistral Medium leads 60.2 to 35.6.
- The biggest single-benchmark swing is MATH Level 5: 36.7% for Llama 3.1-70B and 81.6% for Mistral Medium.
- Llama 3.1-70B is cheaper at $0.40 / $0.40 per million input/output tokens, against $1.50 / $7.50 for Mistral Medium.
- Mistral Medium accepts more context: 262K tokens versus 128K.
Side by side
| Llama 3.1-70B | Mistral Medium | |
|---|---|---|
| Provider | Meta | Mistral AI |
| Noometry Index | 29.4 | 36.0 |
| Released | 2024-07-23 | 2023-12-11 |
| Weights | Open | Open |
| Context window | 128K | 262K |
| Max output | 4K | 262K |
| Input $ / M tokens | $0.40 | $1.50 |
| Output $ / M tokens | $0.40 | $7.50 |
| Results tracked | 35 | 36 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Mistral Medium leads
Llama 3.1-70B: 29.8 (#292), Mistral Medium: 33.8 (#241)
| Benchmark | Llama 3.1-70B | Mistral Medium |
|---|---|---|
| WeirdML | 9% | 43.7% |
| LMArena Coding | 1260 | 1434 |
| FrontierCode | — | 8% |
| SciCode | — | 40.2% |
| BigCodeBench Instruct | 46.1% | — |
| BigCodeBench Complete | 54.8% | — |
| ALE-Bench | — | 763.98 |
Agentic & Tool Use Mistral Medium leads
Llama 3.1-70B: 24.9 (#104), Mistral Medium: 28.1 (#84)
| Benchmark | Llama 3.1-70B | Mistral Medium |
|---|---|---|
| Berkeley Function Calling Leaderboard | — | 37.7% |
| TheAgentCompany | 6.9% | — |
| BALROG | 27.9% | — |
Reasoning Mistral Medium leads
Llama 3.1-70B: 20.8 (#215), Mistral Medium: 23.3 (#165)
| Benchmark | Llama 3.1-70B | Mistral Medium |
|---|---|---|
| LMArena Hard Prompts | 1241 | 1426 |
| DTBench | 60% | 75.5% |
| LMCA | 14.8% | 26.1% |
| Kagi LLM Benchmark | — | 50% |
| CritPt | — | 0% |
| Surface Evolver Bench | — | 26.9% |
| Epoch Capabilities Index | 125.91 | — |
Math Mistral Medium leads
Llama 3.1-70B: 12.7 (#298), Mistral Medium: 27.2 (#238)
| Benchmark | Llama 3.1-70B | Mistral Medium |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 3.6% | 32.2% |
| LMArena Math | 1252 | 1409 |
| MATH Level 5 | 36.7% | 81.6% |
| ProofBench | — | 9% |
| Omni-MATH | 21% | — |
| FrontierMath (Feb 2025 set) | — | 0.3% |
Knowledge Too close to call
Llama 3.1-70B: 24.1 (#264), Mistral Medium: 24.9 (#260)
| Benchmark | Llama 3.1-70B | Mistral Medium |
|---|---|---|
| GPQA Diamond | 44.2% | 59.5% |
| LMArena Expert | 1209 | 1408 |
| Humanity's Last Exam | — | 4.5% |
| MMLU-Pro | 65.3% | — |
| Vectara Hallucination Rate | — | 22.7% |
| GPQA (HELM) | 42.6% | — |
| MMLU | 80.1% | — |
Multimodal Not comparable
Llama 3.1-70B: —, Mistral Medium: 34.6 (#82)
| Benchmark | Llama 3.1-70B | Mistral Medium |
|---|---|---|
| LMArena Vision | — | 1172 |
Multilingual Mistral Medium leads
Llama 3.1-70B: 38.9 (#220), Mistral Medium: 52.3 (#86)
| Benchmark | Llama 3.1-70B | Mistral Medium |
|---|---|---|
| LMArena Non-English | 1219 | 1408 |
| LMArena Chinese | 1215 | 1448 |
| LMArena French | 1260 | 1460 |
| LMArena German | 1222 | 1433 |
| LMArena Japanese | 1132 | 1378 |
| LMArena Korean | 1140 | 1382 |
| LMArena Russian | 1234 | 1411 |
| LMArena Spanish | 1253 | 1433 |
Instruction Following Mistral Medium leads
Llama 3.1-70B: 65.5 (#218), Mistral Medium: 73.9 (#111)
| Benchmark | Llama 3.1-70B | Mistral Medium |
|---|---|---|
| LMArena Instruction Following | 1231 | 1398 |
| IFEval | 82.1% | — |
Long Context Mistral Medium leads
Llama 3.1-70B: 37.6 (#208), Mistral Medium: 42.9 (#108)
| Benchmark | Llama 3.1-70B | Mistral Medium |
|---|---|---|
| LMArena Longer Query | 1241 | 1406 |
Writing & Preference Mistral Medium leads
Llama 3.1-70B: 35.6 (#262), Mistral Medium: 60.2 (#97)
| Benchmark | Llama 3.1-70B | Mistral Medium |
|---|---|---|
| LMArena Text | 1261 | 1424 |
| LMArena Creative Writing | 1232 | 1391 |
| LMArena Multi-Turn | 1256 | 1418 |
| Short-Story Creative Writing | — | 77.3% |
| EQ-Bench Creative Writing | 784 | — |
| WildBench | 75.8% | — |
Frequently asked questions
Is Llama 3.1-70B better than Mistral Medium?
Mistral Medium is the stronger model overall, scoring 36.0 to 29.4 on the Noometry Index. Llama 3.1-70B costs 7.5× less per token, which makes it the better buy when Mistral Medium's lead doesn't matter for your workload.
Which is cheaper, Llama 3.1-70B or Mistral Medium?
Llama 3.1-70B is cheaper. It lists at $0.40 per million input tokens and $0.40 per million output tokens; Mistral Medium lists at $1.50 and $7.50.
Is Llama 3.1-70B or Mistral Medium better for coding?
Mistral Medium scores higher on coding benchmarks: 33.8 versus 29.8 in the Noometry coding category.
Which has the bigger context window?
Mistral Medium does, with 262K tokens against 128K.
How many benchmarks do Llama 3.1-70B and Mistral Medium share?
23 benchmarks have published results for both models. Llama 3.1-70B has 35 scored results on Noometry and Mistral Medium has 36.