Model comparison
Llama-3.3-70B-Instruct vs Mistral Small
Mistral Small is the stronger model overall, scoring 33.4 to 30.6 on the Noometry Index. Llama-3.3-70B-Instruct costs 1.7× less per token, which makes it the better buy when Mistral Small's lead doesn't matter for your workload.
Last verified . 36 shared benchmarks.
Meta
Mistral AISummary
- They share 36 benchmarks with published results for both. Llama-3.3-70B-Instruct scores higher in 1 category and Mistral Small in 8 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in long context, where Mistral Small leads 40.4 to 26.4.
- The biggest single-benchmark swing is LiveBench Instruction Following: 82.7% for Llama-3.3-70B-Instruct and 63.7% for Mistral Small.
- Llama-3.3-70B-Instruct is cheaper at $0.10 / $0.32 per million input/output tokens, against $0.15 / $0.60 for Mistral Small.
- Mistral Small accepts more context: 262K tokens versus 128K.
Side by side
| Llama-3.3-70B-Instruct | Mistral Small | |
|---|---|---|
| Provider | Meta | Mistral AI |
| Noometry Index | 30.6 | 33.4 |
| Released | 2024-12-06 | 2024-02-26 |
| Weights | Open | Open |
| Context window | 128K | 262K |
| Max output | 4K | 256K |
| Input $ / M tokens | $0.10 | $0.15 |
| Output $ / M tokens | $0.32 | $0.60 |
| Results tracked | 43 | 39 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Mistral Small leads
Llama-3.3-70B-Instruct: 31.0 (#290), Mistral Small: 34.0 (#247)
| Benchmark | Llama-3.3-70B-Instruct | Mistral Small |
|---|---|---|
| SciCode | 26% | 26.5% |
| BigCodeBench Instruct | 46.9% | 36.1% |
| LiveBench Coding | 36.6% | 36.2% |
| LMArena Coding | 1268 | 1362 |
| BigCodeBench Complete | 57.5% | 46.6% |
| WeirdML | 14.4% | — |
| ALE-Bench | — | 497.62 |
Agentic & Tool Use Mistral Small leads
Llama-3.3-70B-Instruct: 25.8 (#105), Mistral Small: 28.1 (#93)
| Benchmark | Llama-3.3-70B-Instruct | Mistral Small |
|---|---|---|
| Berkeley Function Calling Leaderboard | 31.9% | 37.1% |
| BALROG | 23% | — |
Reasoning Mistral Small leads
Llama-3.3-70B-Instruct: 14.1 (#327), Mistral Small: 19.8 (#250)
| Benchmark | Llama-3.3-70B-Instruct | Mistral Small |
|---|---|---|
| CritPt | 0% | 0% |
| LiveBench Reasoning | 50.8% | 44.8% |
| LMArena Hard Prompts | 1257 | 1335 |
| DTBench | 59.5% | 70.9% |
| LiveBench Data Analysis | 49.5% | 53.7% |
| LMCA | 17.5% | 20.6% |
| LiveBench | 50.2% | 44% |
| SimpleBench | 19.9% | — |
| Kagi LLM Benchmark | — | 37.8% |
| Epoch Capabilities Index | 127.33 | — |
| ForecastBench | 58.6 | — |
Math Mistral Small leads
Llama-3.3-70B-Instruct: 15.3 (#298), Mistral Small: 16.4 (#293)
| Benchmark | Llama-3.3-70B-Instruct | Mistral Small |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 5.1% | 5.8% |
| LiveBench Math | 42.2% | 39.9% |
| LMArena Math | 1267 | 1341 |
| MATH Level 5 | 41.6% | 46.8% |
Knowledge Too close to call
Llama-3.3-70B-Instruct: 30.6 (#226), Mistral Small: 31.0 (#222)
| Benchmark | Llama-3.3-70B-Instruct | Mistral Small |
|---|---|---|
| GPQA Diamond | 47.4% | 47.5% |
| Vectara Hallucination Rate | 4.1% | 5.1% |
| LMArena Expert | 1225 | 1291 |
| MMLU | 86.3% | 68.7% |
| Confabulations | 22.8% | — |
Multimodal Not comparable
Llama-3.3-70B-Instruct: —, Mistral Small: 33.5 (#96)
| Benchmark | Llama-3.3-70B-Instruct | Mistral Small |
|---|---|---|
| LMArena Vision | — | 1142 |
Multilingual Mistral Small leads
Llama-3.3-70B-Instruct: 39.9 (#220), Mistral Small: 45.5 (#169)
| Benchmark | Llama-3.3-70B-Instruct | Mistral Small |
|---|---|---|
| LMArena Non-English | 1236 | 1315 |
| LMArena Chinese | 1217 | 1340 |
| LMArena French | 1281 | 1337 |
| LMArena German | 1251 | 1340 |
| LMArena Japanese | 1150 | 1275 |
| LMArena Korean | 1143 | 1259 |
| LMArena Russian | 1252 | 1324 |
| LMArena Spanish | 1270 | 1346 |
Instruction Following Llama-3.3-70B-Instruct leads
Llama-3.3-70B-Instruct: 71.1 (#157), Mistral Small: 66.4 (#209)
| Benchmark | Llama-3.3-70B-Instruct | Mistral Small |
|---|---|---|
| LiveBench Instruction Following | 82.7% | 63.7% |
| LMArena Instruction Following | 1242 | 1310 |
Long Context Mistral Small leads
Llama-3.3-70B-Instruct: 26.4 (#295), Mistral Small: 40.4 (#156)
| Benchmark | Llama-3.3-70B-Instruct | Mistral Small |
|---|---|---|
| LMArena Longer Query | 1256 | 1327 |
| Fiction.LiveBench | 33.3% | — |
Writing & Preference Mistral Small leads
Llama-3.3-70B-Instruct: 47.6 (#207), Mistral Small: 52.5 (#171)
| Benchmark | Llama-3.3-70B-Instruct | Mistral Small |
|---|---|---|
| LMArena Text | 1274 | 1338 |
| LMArena Creative Writing | 1250 | 1305 |
| LMArena Multi-Turn | 1280 | 1344 |
| LiveBench Language | 39.2% | 30.5% |
Frequently asked questions
Is Llama-3.3-70B-Instruct better than Mistral Small?
Mistral Small is the stronger model overall, scoring 33.4 to 30.6 on the Noometry Index. Llama-3.3-70B-Instruct costs 1.7× less per token, which makes it the better buy when Mistral Small's lead doesn't matter for your workload.
Which is cheaper, Llama-3.3-70B-Instruct or Mistral Small?
Llama-3.3-70B-Instruct is cheaper. It lists at $0.10 per million input tokens and $0.32 per million output tokens; Mistral Small lists at $0.15 and $0.60.
Is Llama-3.3-70B-Instruct or Mistral Small better for coding?
Mistral Small scores higher on coding benchmarks: 34.0 versus 31.0 in the Noometry coding category.
Which has the bigger context window?
Mistral Small does, with 262K tokens against 128K.
How many benchmarks do Llama-3.3-70B-Instruct and Mistral Small share?
36 benchmarks have published results for both models. Llama-3.3-70B-Instruct has 43 scored results on Noometry and Mistral Small has 39.