Inference
EN

Gemini 3.8 Flash vs GPT-6.1 Terra

Change models
Gemini 3.8 FlashGoogle
GPT-6.1 TerraOpenAI
IndexAll categories, out of 100.
67.8 ±2.5
77.1 ±3.8 best
GapIndex points below the top.
−9.3
Top
CoverageIndex weight that has results.
95% best
90%
SpeedOutput tokens per second.
68/s best
46/s
EloLMArena rating and rank.
1500
—
ReleasedMonth of release.
2026-09
2026-10
Input / 1MUS dollars per 1M input tokens.
$0.825 best
$1.00
Output / 1MUS dollars per 1M output tokens.
$4.13 best
$5.00
Batch output / 1MBatch price per 1M output tokens.
$1.88 best
$2.50
1K chats2K tokens in, 500 out, each.
$3.71 best
$4.50
1K documents20K tokens in, 1K out, each.
$20.63 best
$25.00
1K agent tasks200K tokens in, 20K out, each.
$247.50 best
$300.00
ContextMaximum tokens in one request.
1.05M
1.05M
AgenticMulti-step tasks with tools.
65.3
78.6 best
CodingCode writing and repair.
66.2
74.4 best
ReasoningLogic problems and puzzles.
68.3
81.3 best
MultimodalTasks with images and text.
67.1
73.2 best
KnowledgeFacts and expert knowledge.
70.9 best
70.5
MultilingualTasks in many languages.
—
—
InstructionTasks with strict prompt rules.
91.3
—
MathMath problems.
64.3
76.9 best

Benchmarks

36 shared
Only Gemini 3.8 Flash: 70Only GPT-6.1 Terra: 14
Gemini 3.8 Flash
GPT-6.1 Terra
Artificial Analysis Humanity's Last ExamKnowledge
47.8% best20 / 222
46.2%24 / 222
Artificial Analysis Long Context ReasoningReasoning
81.3%34 / 217
82.7% best23 / 217
GPQA diamond · Epoch AIKnowledge
95.4% best5 / 217
94.3%10 / 217
Critical Physics TasksReasoning
18.3%34 / 214
28.0% best16 / 214
OTIS Mock AIME 2024-2025Math
98.9%22 / 200
100.0% best1 / 200
Artificial Analysis Omniscience AccuracyKnowledge
50.5% best19 / 164
47.6%28 / 164
Vibe Code Bench v1.1Coding
78.7%34 / 123
84.0% best19 / 123
Artificial Analysis SciCodeCoding
56.6% best23 / 114
55.1%28 / 114
Artificial Analysis MMMU-ProMultimodal
85.6% best4 / 113
84.7%6 / 113
SimpleQA VerifiedKnowledge
69.7% best10 / 91
57.3%21 / 91
FrontierMath-Tiers-1-3-v2-PrivateMath
68.4%32 / 90
90.9% best5 / 90
Legal Research BenchLegal
38.9% best38 / 90
26.0%62 / 90
Tax Agent BenchFinance
32.1% best23 / 82
17.1%58 / 82
LiveBench Data AnalysisReasoning
54.0%76 / 80
82.3% best4 / 80
LiveBench ReasoningReasoning
89.3% best18 / 80
88.7%20 / 80
Mystery Game PuzzlesReasoning
47.0%16 / 80
74.0% best4 / 80
FrontierMath-Tier-4-v2-PrivateMath
22.0%48 / 72
92.7% best6 / 72
Terminal-Bench 4.0Agentic
19.7%34 / 61
49.5% best11 / 61
Agent Arena steerabilityAgentic
−1.333 / 54
16.6 best4 / 54
Agent Arena task outcomeAgentic
1.329 / 54
9.7 best11 / 54