BenchLeader
Menu
Best LLMsChoose a modelCoding plansLeaderboardsModelsBenchmarksProvidersReleases4

Claude Sonnet 5 vs GPT-6.1 Terra

Full profiles: Claude Sonnet 5 (Anthropic) and GPT-6.1 Terra (OpenAI): every benchmark, price and rank.

Updated · Data from 8 sources

As of 2 Nov 2026, GPT-6.1 Terra is ahead of Claude Sonnet 5 on quality (65.2 vs 59.4 on the BenchLeader Index), and GPT-6.1 Terra is 2.0× cheaper at $2.00 vs $4.00 per 1M tokens blended. GPT-6.1 Terra leads in every category we measure both in. GPT-6.1 Terra streams faster.

Key facts (as of 2 Nov 2026)

  • Claude Sonnet 5 scores 59.4 on the BenchLeader Index and GPT-6.1 Terra scores 65.2 (BenchLeader, 2 Nov 2026).
  • Claude Sonnet 5 costs $4.00 per 1M tokens blended (list price, 2 Nov 2026).
  • GPT-6.1 Terra costs $2.00 per 1M tokens blended (list price, 2 Nov 2026).
  • GPT-6.1 Terra streams faster than Claude Sonnet 5 (Artificial Analysis, 7-day median, 2 Nov 2026).
  • Of the 12 benchmarks both models have, GPT-6.1 Terra leads 12 (BenchLeader, 2 Nov 2026).
  • On Arena WebDev, Claude Sonnet 5 is rated 1540 and GPT-6.1 Terra 1673 (LMArena, as published by 2 Nov 2026).
Claude Sonnet 5 benchmarks high Anthropic GPT-6.1 Terra benchmarks max OpenAIAdd a third model

Which to pick for…

The pick's numbers come first. Category scores on the BenchLeader scale (50 is the average model, 15 points one standard deviation). High confidence means a lead of half a standard deviation or more; under one point is called either.

Use casePickWhyConfidence
CodingGPT-6.1 Terra70 vs 60 category score; Arena WebDev 1673 vs 1540High
Agents & tool useGPT-6.1 Terra69 vs 55 category score; Arena Agent 10.6 vs 2.9High
MathsGPT-6.1 Terra73 vs 65 category scoreMedium
ReasoningGPT-6.1 Terra70 vs 66 category score; Humanity's Last Exam (AA) 46.3% vs 35.7%Medium
Long documentsGPT-6.1 Terra66 vs 63 category score; AA-LCR v1.1 82.7% vs 76.7%Medium
Images and other mediaGPT-6.1 Terra67 vs 61 category scoreMedium
Chat quality (human preference)–Only Claude Sonnet 5 has an Arena rating (65 category score)Not comparable
Lowest costGPT-6.1 Terra$2.00 vs $4.00 per 1M blended; agentic coding session $0.080 vs $0.160–
Fastest first answerClaude Sonnet 510.3 s (Claude Sonnet 5 (high)) vs 264.3 s (GPT-6.1 Terra (max)) to the first answer token–
High volume on a budgetNeither flagshipA smaller model is usually the better fit: Cheapest good LLMs, Fastest LLMs–

Key differences

  • GPT-6.1 Terra (max) leads on quality: 65.2 vs 59.4.
  • Claude Sonnet 5 (high) is stronger in human preference.
  • GPT-6.1 Terra (max) is stronger in agents & tools, coding, long context, maths, multimodal, reasoning, knowledge.
  • GPT-6.1 Terra (max) is 2.0× cheaper ($2.00 vs $4.00 per 1M blended).
  • GPT-6.1 Terra (max) streams 1.5× faster (107 vs 71 tokens per second).

At a glance

SpecClaude Sonnet 5GPT-6.1 Terra
DeveloperAnthropicOpenAI
Released30 Jun 202626 Oct 2026
Context window (tokens)1,000,0001,050,000
Compared atClaude Sonnet 5 (high)GPT-6.1 Terra (max)
Open weightsNoNo
Price in / out / cached input per 1M$2.00 / $10.00 / $0.200$1.00 / $5.00 / $0.050

Category by category

Category scores on the BenchLeader scale: 50 is the average evaluated model, 15 points is one standard deviation. A gap under one point is called level.

CategoryClaude Sonnet 5 (high)GPT-6.1 Terra (max)Leader
Reasoning65.569.6GPT-6.1 Terra
Coding60.470.0GPT-6.1 Terra
Agents & tools55.368.6GPT-6.1 Terra
Maths65.472.7GPT-6.1 Terra
Knowledge–62.8not comparable
Human preference64.8–not comparable
Multimodal61.067.4GPT-6.1 Terra
Long context63.166.2GPT-6.1 Terra

On one chart

Each domain on the index's own scale, where 50 is an average model. Click a domain to see the benchmarks behind it. Pick up to 5 models from the list.

2550 avg75ReasoningCodingAgents & toolsMathsLong contextMultimodalClaude Sonnet 5, Reasoning: 65.5Claude Sonnet 5, Coding: 60.4Claude Sonnet 5, Agents & tools: 55.3Claude Sonnet 5, Maths: 65.4Claude Sonnet 5, Long context: 63.1Claude Sonnet 5, Multimodal: 61.0GPT-6.1 Terra, Reasoning: 69.6GPT-6.1 Terra, Coding: 70.0GPT-6.1 Terra, Agents & tools: 68.6GPT-6.1 Terra, Maths: 72.7GPT-6.1 Terra, Long context: 66.2GPT-6.1 Terra, Multimodal: 67.4

Not drawn, measured for fewer than two of these models: Knowledge, Instruction following, Human preference.

ModelIndex

Where sources disagree

Across the 12 benchmarks both have, Claude Sonnet 5 leads 0, GPT-6.1 Terra leads 12. Every overall index that rates both puts GPT-6.1 Terra ahead.

Overall indexClaude Sonnet 5 (high)GPT-6.1 Terra (max)Leader
BenchLeader Index59.465.2GPT-6.1 Terra

Each publisher runs its own tasks, harnesses and effort settings, so their overall verdicts can differ; BenchLeader reconciles them on one scale. † measured at another effort setting of the same model (the publisher did not test the one compared here).

Price, speed and latency

List prices blended at three input tokens per output token. Per-task prices use the token counts on the methodology page. Cost, tokens and time per benchmark task are measured: Artificial Analysis’ averages over its whole Intelligence Index run.

MeasureClaude Sonnet 5 (high)GPT-6.1 Terra (max)
Blended price per 1M tokens$4.00$2.00
Input / output per 1M$2.00 / $10.00$1.00 / $5.00
Cached input per 1M$0.200$0.050
Price per task: chat reply$0.0038$0.0019
Price per task: summarize a 30-page report$0.030$0.015
Price per task: code edit$0.027$0.013
Price per task: agentic coding session$0.160$0.080
Price per task: structured extraction$0.0060$0.0030
Agentic coding session, input mostly cached$0.079$0.037
Cost per benchmark task (AA Index)–$0.961
Output tokens per task (AA Index)–60k
Of which reasoning–39k
Time per task (AA Index)–9 min
Output speed71 tok/s107 tok/s
Time to first answer10.3 s264.3 s
Time to a complete answer17.3 s269.0 s

Each model is compared at its best-ranked reasoning-effort setting. Claude Sonnet 5 (high) scores 59.4 and takes 10 s to its first answer token; Claude Sonnet 5 scores 55.9 in 2 s, Claude Sonnet 5 (xhigh) scores 55.6 in 25 s. GPT-6.1 Terra (max) scores 65.2 and takes 264 s to its first answer token; GPT-6.1 Terra (xhigh) scores 60.6 in 69 s, GPT-6.1 Terra (high) scores 60.3 in 21 s. Higher effort means more thinking before the answer: better scores, longer waits and more tokens billed. How reasoning effort changes scores and cost.

Every benchmark

124 benchmarks, grouped by category; each name links to the board and its leaders. † measured at another effort setting of the same model, shown for reference and not counted as a win.

MetricPublisherClaude Sonnet 5 (high)GPT-6.1 Terra (max)
BenchLeader IndexBenchLeader59.465.2
Agents & tools scoreBenchLeader55.368.6
Coding scoreBenchLeader60.470.0
Human preference scoreBenchLeader64.8–
Long context scoreBenchLeader63.166.2
Maths scoreBenchLeader65.472.7
Multimodal scoreBenchLeader61.067.4
Reasoning scoreBenchLeader65.569.6
Knowledge scoreBenchLeader–62.8
Blended price $/MBenchLeader$4.00$2.00
Output speedBenchLeader71 tok/s107 tok/s
Time to first answerBenchLeader10.3 s264.3 s
Context windowBenchLeader1,000,000 tokens1,050,000 tokens
Reasoning
GPQA DiamondEpoch AI Benchmarking Hub90.5%†94.3%
SimpleBenchSimpleBench60.6%†–
Arena Hard PromptsArena (LMArena)1491–
LiveBench 2.0 ReasoningLiveBench–88.7%†
LiveBench 1.0 ReasoningLiveBench88.7%†–
GPQA Diamond (AA)Artificial Analysis91.1%†–
Humanity's Last Exam (AA)Artificial Analysis35.7%46.3%
HLE-Diamond (AA)Artificial Analysis–47.3%
GPQA Diamond (Vals)Vals AI88.9%†–
CritPtArtificial Analysis15.1%28.0%
CritPt v2 hard subset (AA)Artificial Analysis–64.5%
Chess PuzzlesEpoch AI Benchmarking Hub34.0%†–
Mystery Game PuzzlesEpoch AI Benchmarking Hub16.0%†74.0%
LMCAEpoch AI Benchmarking Hub50.0%–
DTBenchEpoch AI Benchmarking Hub84.5%–
ForecastBenchForecastBench63.5%†–
BTF-3Epoch AI Benchmarking Hub14.2%†–
Coding
SciCodeSciCode54.3%55.1%
WeirdMLWeirdML68.8%–
FrontierCodeCognition42.7%†–
GSO-BenchGSO-Bench37.3%†–
Arena CodingArena (LMArena)1520–
Arena WebDevArena (LMArena)15401673
LiveBench 2.0 CodingLiveBench–94.3%†
LiveBench 1.0 CodingLiveBench80.7%†–
Show all 124 benchmarks
MetricPublisherClaude Sonnet 5 (high)GPT-6.1 Terra (max)
Coding
LiveCodeBenchVals AI82.4%†–
IOIVals AI45.0%†–
SWE-bench (Vals)Vals AI79.6%†–
Code MigrationVals AI44.4%†–
ProgramBenchVals AI0.0%†–
Vibe Code Bench 1-100Vals AI13.8%†–
Vibe Code Bench v1.1Vals AI81.3%†84.0%
Surface Evolver BenchEpoch AI Benchmarking Hub60.0%†–
DeepSWE v1.1Epoch AI Benchmarking Hub48.2%–
CursorBenchCursor30.8%–
ALE-BenchEpoch AI Benchmarking Hub1463.12608.8
Terminal-Bench 4.0 (AA)Artificial Analysis5.0%55.6%
Terminal-Bench 2.1 (AA)Artificial Analysis80.5%†–
Startrise FrontendStartrise66.6%†82.0%†
GameBench 2 OverallDuelLab GameBench 248.1†–
GameBench 2 by reasoning settingDuelLab GameBench 256.5†–
FrontierCode 1.1 (AA)Artificial Analysis–36.0%
FrontierCode 2.0 (AA)Artificial Analysis–42.3%
Agents & tools
Terminal-BenchTerminal-Bench12.4%†–
APEX-AgentsMercor54.5%†68.3%
Arena AgentArena (LMArena)2.910.6
LiveBench Long AgenticLiveBench–48.6%†
LiveBench BrowserLiveBench–93.1%†
LiveBench Agentic CodingLiveBench59.4%†–
Terminal-Bench 2.1 (Vals)Vals AI74.5%†–
GDPval-AA v2.1Artificial Analysis–1476
τ³-Banking (AA)Artificial Analysis37.3%†–
Analyst Agent (AA)Artificial Analysis46.3%†–
CyberBenchVals AI61.9%†–
Finance Agent v2Vals AI53.9%†–
Finance Agent v2.1Vals AI–56.0%
Harvey's Legal Agent Benchmark v1.1Vals AI–1.7%
Harvey's Legal Agent Benchmark v1Vals AI5.0%†–
Legal Research BenchVals AI41.8%†26.0%
SkillsBenchVals AI46.5%†–
SREBenchVals AI0.4%†–
Tax Agent BenchVals AI29.4%†17.1%
Terminal-Bench 4.0 (Vals)Vals AI10.6%†49.5%
Terminal-Bench ScienceVals AI5.7%†–
GBAEvalEpoch AI Benchmarking Hub65.3%†–
Vending-Bench 2Andon Labs6377.7†9107.5†
AutomationBenchArtificial Analysis32.1%60.2%
GDP.pdf (AA)Artificial Analysis9.4%28.6%
EnterpriseOps-GymArtificial Analysis43.9%†–
AA-Briefcase v1.1Artificial Analysis1357†1402
GDPxGDPx by SemiAnalysis–8.5%
Terminal-Bench-Science 0.1 (AA)Artificial Analysis–57.6%
Maths
FrontierMath Tiers 1–3Epoch AI Benchmarking Hub65.6%†90.9%
FrontierMath Tier 4Epoch AI Benchmarking Hub29.3%†92.7%
OTIS Mock AIMEEpoch AI Benchmarking Hub94.7%†100.0%
ProofBenchVals AI77.0%†–
LiveBench MathematicsLiveBench92.9%†–
Arena MathsArena (LMArena)1475–
Knowledge
SimpleQA VerifiedEpoch AI Benchmarking Hub32.9%†57.3%
LiveBench 2.0 Data AnalysisLiveBench–82.3%†
LiveBench ResearchLiveBench–73.9%†
LiveBench 1.0 Data AnalysisLiveBench71.7%†–
AA-Omniscience v1.1Artificial Analysis–29.6
MMLU-ProVals AI87.5%†–
LegalBenchVals AI83.9%†–
CorpFinVals AI67.0%†–
TaxEvalVals AI75.6%†–
Excel Modeling BenchmarkVals AI66.3%†–
Excel Modeling Benchmark v1.1Vals AI–73.3%
MedCodeVals AI47.5%†–
MedScribeVals AI76.0%†–
Public Benefits Bench v1.1Vals AI66.0%†–
DrugDiscoveryBenchScale AI SEAL50.0%†–
EBR-benchEpoch AI Benchmarking Hub–78.1%
AA-Omniscience v1.1: accuracyArtificial Analysis–47.6%
AA-Omniscience v1.1: non-hallucinationArtificial Analysis–65.5%
Instruction following
LiveBench LanguageLiveBench75.0%†–
LiveBench Instruction FollowingLiveBench63.9%†–
MultiChallengeScale AI SEAL–74.2%†
Arena Instruction FollowingArena (LMArena)1467–
Human preference
Arena TextArena (LMArena)1462–
EQ-Bench 4EQ-Bench1221†–
Arena Creative WritingArena (LMArena)1436–
Arena Multi-turnArena (LMArena)1472–
Arena Longer QueriesArena (LMArena)1481–
Multimodal
Arena VisionArena (LMArena)1273–
LiveBench DesignLiveBench–82.3%†
MMMU-ProArtificial Analysis77.3%†84.7%
MMMU-Pro (Vals)Vals AI83.0%†–
MortgageTaxVals AI70.0%†–
Blueprint-Bench 2Epoch AI Benchmarking Hub24.9%†42.1%†
Long context
AA-LCR v1.1Artificial Analysis76.7%82.7%
Arena DocumentArena (LMArena)1476–
MLCRArtificial Analysis4.4%30.6%
Composite
Epoch Capabilities IndexEpoch AI Benchmarking Hub157.0†168.1†
LiveBench 2.0LiveBench–77.8%†
LiveBench 1.0LiveBench76.0%†–
AA Intelligence Index v5.0Artificial Analysis–51.0
AA Intelligence Index v4.3.2Artificial Analysis31.748.1
AA Progress Index v1Artificial Analysis296.8345.2
Vals IndexVals AI–60.5
Vals Finance IndexVals AI–53.3%
Vals Multimodal IndexVals AI68.8%†–
Safety & honesty
FORTRESSScale AI SEAL–18.2%†

Data as of 2026-11-02. Best configuration of each model; every score links to its source on the model pages.

DuelLab GameBench 2, release gb2-20261026-cb69502a12d9 (https://benchmarks.duellab.org/releases/gb2-20261026-cb69502a12d9/), CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). Scores mapped to BenchLeader model names; not endorsed by DuelLab.

What changed

  • GPT-6.1 Terra (max): 1 score revision.
  • GPT-6.1 Terra (max): BenchLeader Index 64.7 → 65.2.
  • GPT-6.1 Terra (max): 1 new result (Arena Agent).
  • Claude Sonnet 5 (high): 1 score revision.
  • GPT-6.1 Terra (max): 3 new results (SciCode, APEX-Agents, ALE-Bench).
  • GPT-6.1 Terra (max): BenchLeader Index 63.5 → 64.7.
  • Claude Sonnet 5 (high): 1 new result (AA Progress Index v1).
  • GPT-6.1 Terra (max): 7 new results (AA Intelligence Index v5.0, AA Progress Index v1, HLE-Diamond (AA)…); 2 score revisions; cost per task moved from $0.40/M to $0.96/M (138% up).
  • Claude Sonnet 5 (high): BenchLeader Index 60.0 → 59.5.
  • Claude Sonnet 5 (high): 1 new result (MLCR).

Rankings: Best LLM overall · Best LLM for coding · Best LLM for agents · Best cheap LLMs

Claude Sonnet 5 vs GPT-6.1 Terra: questions

Is GPT-6.1 Terra better than Claude Sonnet 5?
GPT-6.1 Terra (max) leads on quality: 65.2 vs 59.4. The BenchLeader Index combines every independent quality benchmark; GPT-6.1 Terra (max) is ahead overall as of 2 Nov 2026, but check the category scores for your use.
Claude Sonnet 5 vs GPT-6.1 Terra: which is cheaper, and will that change?
GPT-6.1 Terra is cheaper today: $2.00 against $4.00 per million tokens, blended at three input tokens per output token. No price change has been announced for either.
Is Claude Sonnet 5 or GPT-6.1 Terra better for coding?
GPT-6.1 Terra scores higher in coding (70 vs 60 on the category index, where 50 is average). On Arena WebDev, the top shared coding benchmark, it is 1540 for Claude Sonnet 5 against 1673 for GPT-6.1 Terra.
Which is faster, Claude Sonnet 5 or GPT-6.1 Terra?
GPT-6.1 Terra streams faster (107 vs 71 output tokens per second, Artificial Analysis 7-day median).
Is Claude Sonnet 5 or GPT-6.1 Terra better for agents and tool use?
GPT-6.1 Terra scores higher in agentic tasks (69 vs 55 on the category index, where 50 is average). On Arena Agent, the top shared agents & tools benchmark, it is 2.9 for Claude Sonnet 5 against 10.6 for GPT-6.1 Terra.
More questions (4)
Which is better for long documents?
GPT-6.1 Terra scores higher on long-context benchmarks (66 vs 63). Both take about 1.1M tokens: Claude Sonnet 5 1,000,000, GPT-6.1 Terra 1,050,000.
Which has lower latency?
Claude Sonnet 5 answers first: 10.3 s against 264.3 s to the first answer token, as Artificial Analysis measures it. Reasoning effort drives most of this: a higher setting thinks longer before it answers.
Do independent benchmarks agree on Claude Sonnet 5 vs GPT-6.1 Terra?
Across the 12 benchmarks both have, Claude Sonnet 5 leads 0, GPT-6.1 Terra leads 12. Every overall index that rates both puts GPT-6.1 Terra ahead.
When was GPT-6.1 Terra released?
GPT-6.1 Terra was released on 26 Oct 2026.