GPT-5 vs GPT-OSS 20B

Change models
GPT-5OpenAI
IndexAll categories, out of 100.
53.5 ±3.4 best
41.6 ±4.6
GapIndex points below the top.
Top
−11.9
CoverageIndex weight that has results.
95% best
90%
SpeedOutput tokens per second.
77/s
496/s best
EloLMArena rating and rank.
1406 best
1288
ReleasedMonth of release.
2025-08
2025-08
Input / 1MUS dollars per 1M input tokens.
$1.25
$0.017 best
Output / 1MUS dollars per 1M output tokens.
$10.00
$0.085 best
Batch output / 1MBatch price per 1M output tokens.
$5.00
—
1K chats2K tokens in, 500 out, each.
$7.50
$0.077 best
1K documents20K tokens in, 1K out, each.
$35.00
$0.425 best
1K agent tasks200K tokens in, 20K out, each.
$450.00
$5.10 best
ContextMaximum tokens in one request.
400K best
131K
AgenticMulti-step tasks with tools.
49.8 best
40.0
CodingCode writing and repair.
51.3 best
48.5
ReasoningLogic problems and puzzles.
54.3 best
35.2
MultimodalTasks with images and text.
55.9
—
KnowledgeFacts and expert knowledge.
57.4 best
35.8
MultilingualTasks in many languages.
—
37.2
InstructionTasks with strict prompt rules.
56.5 best
49.3
MathMath problems.
54.8 best
45.7

Benchmarks

20 shared
Only GPT-5: 40Only GPT-OSS 20B: 17
GPT-5
GPT-OSS 20B
86.2% best72 / 220
46.0%164 / 220
28.5% best84 / 216
11.0%135 / 216
78.2% best68 / 211
34.7%169 / 211
5.7% best77 / 207
1.4%108 / 207
91.4% best59 / 203
50.8%125 / 203
85.4% best73 / 193
68.8%128 / 193
86.0% best17 / 149
70.8%122 / 149
TaxEval v2Finance
73.4% best49 / 143
63.7%119 / 143
84.8% best55 / 143
60.2%81 / 143
85.9% best25 / 142
80.4%69 / 142
73.1% best31 / 140
65.1%52 / 140
85.6% best45 / 135
68.9%97 / 135
MMLU ProKnowledge
86.5% best41 / 135
71.6%115 / 135
20.9% best81 / 132
0.0%100 / 132
CorpFin v2Finance
61.1% best58 / 131
53.1%99 / 131
AIMEMath
93.4% best14 / 95
86.0%33 / 95
MedQAHealthcare
96.3% best4 / 95
82.9%65 / 95
MGSMMultilingual
92.8% best14 / 74
89.0%50 / 74
96.0% best3 / 59
94.2%10 / 59
66.5% best6 / 53
43.8%53 / 53