Model comparison
gpt-oss-20b vs Mixtral 8x22B
gpt-oss-20b is the stronger model overall, scoring 32.5 to 27.1 on the Noometry Index.
Last verified . 25 shared benchmarks.
Mistral AISummary
- They share 25 benchmarks with published results for both. gpt-oss-20b scores higher in 6 categories and Mixtral 8x22B in 3 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where gpt-oss-20b leads 34.6 to 15.1.
- The biggest single-benchmark swing is Omni-MATH: 56.5% for gpt-oss-20b and 16.3% for Mixtral 8x22B.
- gpt-oss-20b is cheaper at $0.018 / $0.09 per million input/output tokens, against $2 / $6 for Mixtral 8x22B.
- gpt-oss-20b accepts more context: 131K tokens versus 64K.
Side by side
| gpt-oss-20b | Mixtral 8x22B | |
|---|---|---|
| Provider | OpenAI | Mistral AI |
| Noometry Index | 32.5 | 27.1 |
| Released | 2025-08-05 | 2024-04-17 |
| Weights | Open | Open |
| Context window | 131K | 64K |
| Max output | 16K | 64K |
| Input $ / M tokens | $0.018 | $2 |
| Output $ / M tokens | $0.09 | $6 |
| Results tracked | 34 | 34 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding gpt-oss-20b leads
gpt-oss-20b: 37.6 (#192), Mixtral 8x22B: 24.2 (#329)
| Benchmark | gpt-oss-20b | Mixtral 8x22B |
|---|---|---|
| WeirdML | 40.9% | 3.2% |
| LMArena Coding | 1306 | 1166 |
| SciCode | 34.4% | — |
| BigCodeBench Instruct | — | 40.6% |
| BigCodeBench Complete | — | 50.2% |
| ALE-Bench | 566.05 | — |
| HumanEval+ | — | 72% |
| MBPP+ | — | 64.3% |
Agentic & Tool Use Mixtral 8x22B leads
gpt-oss-20b: 9.3 (#154), Mixtral 8x22B: 23.1 (#127)
| Benchmark | gpt-oss-20b | Mixtral 8x22B |
|---|---|---|
| Terminal-Bench | 3.4% | — |
| Cybench | — | 7.5% |
Reasoning Too close to call
gpt-oss-20b: 19.3 (#261), Mixtral 8x22B: 19.9 (#248)
| Benchmark | gpt-oss-20b | Mixtral 8x22B |
|---|---|---|
| LMArena Hard Prompts | 1274 | 1150 |
| DTBench | 68% | 55.1% |
| Epoch Capabilities Index | 137.82 | 122.03 |
| Kagi LLM Benchmark | 53.2% | — |
| CritPt | 1.4% | — |
| Chess Puzzles | 4% | — |
| LMCA | 14.5% | — |
| ForecastBench | — | 56.3 |
Math gpt-oss-20b leads
gpt-oss-20b: 39.4 (#103), Mixtral 8x22B: 22.9 (#275)
| Benchmark | gpt-oss-20b | Mixtral 8x22B |
|---|---|---|
| Omni-MATH | 56.5% | 16.3% |
| LMArena Math | 1317 | 1184 |
| OTIS Mock AIME 2024-2025 | 65.3% | — |
| MATH Level 5 | — | 24.2% |
Knowledge gpt-oss-20b leads
gpt-oss-20b: 34.6 (#195), Mixtral 8x22B: 15.1 (#293)
| Benchmark | gpt-oss-20b | Mixtral 8x22B |
|---|---|---|
| GPQA Diamond | 60.8% | 34.1% |
| MMLU-Pro | 74% | 46% |
| GPQA (HELM) | 59.4% | 33.4% |
| LMArena Expert | 1258 | 1113 |
| MMLU | — | 77.8% |
Multilingual gpt-oss-20b leads
gpt-oss-20b: 42.2 (#197), Mixtral 8x22B: 32.8 (#255)
| Benchmark | gpt-oss-20b | Mixtral 8x22B |
|---|---|---|
| LMArena Non-English | 1268 | 1128 |
| LMArena Chinese | 1314 | 1116 |
| LMArena German | 1255 | 1141 |
| LMArena Japanese | 1244 | 1037 |
| LMArena Korean | 1236 | 1057 |
| LMArena Russian | 1278 | 1158 |
| LMArena Spanish | 1267 | 1151 |
| LMArena French | — | 1166 |
Instruction Following gpt-oss-20b leads
gpt-oss-20b: 61.8 (#240), Mixtral 8x22B: 57.7 (#266)
| Benchmark | gpt-oss-20b | Mixtral 8x22B |
|---|---|---|
| IFEval | 73.2% | 72.4% |
| LMArena Instruction Following | 1236 | 1147 |
Long Context gpt-oss-20b leads
gpt-oss-20b: 37.9 (#209), Mixtral 8x22B: 34.7 (#247)
| Benchmark | gpt-oss-20b | Mixtral 8x22B |
|---|---|---|
| LMArena Longer Query | 1250 | 1144 |
Writing & Preference Mixtral 8x22B leads
gpt-oss-20b: 35.5 (#265), Mixtral 8x22B: 36.9 (#262)
| Benchmark | gpt-oss-20b | Mixtral 8x22B |
|---|---|---|
| LMArena Text | 1287 | 1162 |
| LMArena Creative Writing | 1201 | 1141 |
| WildBench | 73.7% | 71.1% |
| LMArena Multi-Turn | 1268 | 1130 |
| EQ-Bench Creative Writing | 666 | — |
Frequently asked questions
Is gpt-oss-20b better than Mixtral 8x22B?
gpt-oss-20b is the stronger model overall, scoring 32.5 to 27.1 on the Noometry Index.
Which is cheaper, gpt-oss-20b or Mixtral 8x22B?
gpt-oss-20b is cheaper. It lists at $0.018 per million input tokens and $0.09 per million output tokens; Mixtral 8x22B lists at $2 and $6.
Is gpt-oss-20b or Mixtral 8x22B better for coding?
gpt-oss-20b scores higher on coding benchmarks: 37.6 versus 24.2 in the Noometry coding category.
Which has the bigger context window?
gpt-oss-20b does, with 131K tokens against 64K.
How many benchmarks do gpt-oss-20b and Mixtral 8x22B share?
25 benchmarks have published results for both models. gpt-oss-20b has 34 scored results on Noometry and Mixtral 8x22B has 34.