Model comparison
Ministral 8B vs Qwen3 14B
Qwen3 14B is the stronger model overall, scoring 35.5 to 28.2 on the Noometry Index. Ministral 8B costs 4.1× less per token, which makes it the better buy when Qwen3 14B's lead doesn't matter for your workload.
Last verified . 4 shared benchmarks.
Mistral AI
Alibaba (Qwen)Summary
- They share 4 benchmarks with published results for both. Ministral 8B scores higher in 0 categories and Qwen3 14B in 6 categories; 5 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where Qwen3 14B leads 39.3 to 12.6.
- The biggest single-benchmark swing is GPQA Diamond: 27.1% for Ministral 8B and 63.8% for Qwen3 14B.
- Ministral 8B is cheaper at $0.15 / $0.15 per million input/output tokens, against $0.35 / $1.40 for Qwen3 14B.
- Ministral 8B accepts more context: 262K tokens versus 131K.
Side by side
| Ministral 8B | Qwen3 14B | |
|---|---|---|
| Provider | Mistral AI | Alibaba (Qwen) |
| Noometry Index | 28.2 | 35.5 |
| Released | 2024-10-01 | 2025-04 |
| Weights | Open | Open |
| Context window | 262K | 131K |
| Max output | 262K | 8K |
| Input $ / M tokens | $0.15 | $0.35 |
| Output $ / M tokens | $0.15 | $1.40 |
| Results tracked | 17 | 12 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Qwen3 14B leads
Ministral 8B: 35.0 (#230), Qwen3 14B: 37.3 (#195)
| Benchmark | Ministral 8B | Qwen3 14B |
|---|---|---|
| SciCode | — | 31.6% |
| LMArena Coding | 1202 | — |
Agentic & Tool Use Qwen3 14B leads
Ministral 8B: 16.4 (#148), Qwen3 14B: 29.6 (#83)
| Benchmark | Ministral 8B | Qwen3 14B |
|---|---|---|
| Berkeley Function Calling Leaderboard | 11.1% | 41% |
Reasoning Too close to call
Ministral 8B: 18.4 (#281), Qwen3 14B: 18.5 (#280)
| Benchmark | Ministral 8B | Qwen3 14B |
|---|---|---|
| DTBench | 45.7% | 64% |
| Kagi LLM Benchmark | — | 49.1% |
| CritPt | — | 0% |
| Chess Puzzles | — | 4% |
| LMArena Hard Prompts | 1191 | — |
| LMCA | — | 18.2% |
| Epoch Capabilities Index | — | 138.23 |
Math Qwen3 14B leads
Ministral 8B: 25.7 (#267), Qwen3 14B: 38.6 (#133)
| Benchmark | Ministral 8B | Qwen3 14B |
|---|---|---|
| OTIS Mock AIME 2024-2025 | — | 66.4% |
| LMArena Math | 1188 | — |
| MATH Level 5 | 14.9% | — |
Knowledge Qwen3 14B leads
Ministral 8B: 12.6 (#297), Qwen3 14B: 39.3 (#134)
| Benchmark | Ministral 8B | Qwen3 14B |
|---|---|---|
| GPQA Diamond | 27.1% | 63.8% |
| Vectara Hallucination Rate | 7.4% | 5.4% |
| LMArena Expert | 1170 | — |
Multilingual Not comparable
Ministral 8B: 35.1 (#247), Qwen3 14B: —
| Benchmark | Ministral 8B | Qwen3 14B |
|---|---|---|
| LMArena Non-English | 1165 | — |
| LMArena Chinese | 1193 | — |
| LMArena Russian | 1195 | — |
Instruction Following Not comparable
Ministral 8B: 60.5 (#250), Qwen3 14B: —
| Benchmark | Ministral 8B | Qwen3 14B |
|---|---|---|
| LMArena Instruction Following | 1161 | — |
Long Context Qwen3 14B leads
Ministral 8B: 36.7 (#227), Qwen3 14B: 38.1 (#204)
| Benchmark | Ministral 8B | Qwen3 14B |
|---|---|---|
| Fiction.LiveBench | — | 62.5% |
| LMArena Longer Query | 1212 | — |
Writing & Preference Not comparable
Ministral 8B: 39.6 (#246), Qwen3 14B: —
| Benchmark | Ministral 8B | Qwen3 14B |
|---|---|---|
| LMArena Text | 1191 | — |
| LMArena Creative Writing | 1175 | — |
| LMArena Multi-Turn | 1166 | — |
Frequently asked questions
Is Ministral 8B better than Qwen3 14B?
Qwen3 14B is the stronger model overall, scoring 35.5 to 28.2 on the Noometry Index. Ministral 8B costs 4.1× less per token, which makes it the better buy when Qwen3 14B's lead doesn't matter for your workload.
Which is cheaper, Ministral 8B or Qwen3 14B?
Ministral 8B is cheaper. It lists at $0.15 per million input tokens and $0.15 per million output tokens; Qwen3 14B lists at $0.35 and $1.40.
Is Ministral 8B or Qwen3 14B better for coding?
Qwen3 14B scores higher on coding benchmarks: 37.3 versus 35.0 in the Noometry coding category.
Which has the bigger context window?
Ministral 8B does, with 262K tokens against 131K.
How many benchmarks do Ministral 8B and Qwen3 14B share?
4 benchmarks have published results for both models. Ministral 8B has 17 scored results on Noometry and Qwen3 14B has 12.