Head-to-head comparison

GPT-5.6 Terra vs. GPT-5.5

Compare published benchmark results from matching versions and cohorts, API pricing, and technical specifications. Any documented test setup differences remain visible.

shared result series: 238
distinct benchmarks: 27
latest retrieval: 2026-09-04

First modelSecond model

Specifications

General specifications

Release date, availability, context window, architecture, and published parameter counts at a glance. Undisclosed details remain marked as unavailable.

Specifications for GPT-5.6 Terra and GPT-5.5
Attribute	GPT-5.6 Terra	GPT-5.5
Provider	OpenAI	OpenAI
Released	Jun 26, 2026	Apr 2026
Availability	Active	Active
Model type	Proprietary	Proprietary
Parameters	Not published	Not published
Architecture	Not published	Not published
Context window	1,050,000 tokens	1,050,000 tokens
Knowledge cutoff	February 16, 2026	December 1, 2025
Technical sources	Model, Context, Knowledge cutoff	Model

Pricing

API pricing compared

The table compares official standard API rates per 1M tokens, including published cache pricing. Batch, fast, priority, regional surcharges, tool calls, and cloud platform pricing are excluded.

API pricing for GPT-5.6 Terra and GPT-5.5
Price type	GPT-5.6 Terra	GPT-5.5
API input	$2 ≤272K / $4 >272K	$5 ≤272K / $10 >272K
API output	$12 ≤272K / $18 >272K	$30 ≤272K / $45 >272K
Cache read	$0.2 ≤272K / $0.4 >272K	$0.5 ≤272K / $1 >272K
Cache write (5 min.)	$2.5 ≤272K / $5 >272K	Not listed
Cache write (1 hr.)	Not listed	Not listed
Price verified	08/11/2026 Source	07/30/2026 Source

GPT-5.6 TerraGPT-5.5

API inputUSD per 1M tokens, base tier

GPT-5.6 Terra$2

GPT-5.5$5

API outputUSD per 1M tokens, base tier

GPT-5.6 Terra$12

GPT-5.5$30

Cache readUSD per 1M tokens, base tier

GPT-5.6 Terra$0.2

GPT-5.5$0.5

Key to the marks:better valuebetter, but not conclusivetie

Pricing by provider

Direct developer pricing and provider endpoint tariffs routed through OpenRouter are listed separately. Prices are in USD per 1M tokens. Context-dependent direct tiers plus regional, Flex, Priority, and other OpenRouter tariffs remain individually identifiable.

Provider pricing for GPT-5.6 Terra and GPT-5.5
Model	Provider	Purchase route and tariff	Input per 1M tokens	Output per 1M tokens	Source
GPT-5.6 Terra	OpenAI	Direct from developerUp to 272,000 context tokens	$2	$12	Source08/11/2026
	OpenAI	Direct from developerAbove 272,000 context tokens	$4	$18	Source08/11/2026
	Amazon Bedrock	Via OpenRouteramazon-bedrock/us-east-1	$2.2	$13.2	Source09/07/2026
	Azure	Via OpenRouterazure	$2	$12	Source09/07/2026
	Azure	Via OpenRouterazure/eu	$2.2	$13.2	Source09/07/2026
	Azure	Via OpenRouterazure/us	$2.2	$13.2	Source09/07/2026
	OpenAI	Via OpenRouteropenai	$2	$12	Source09/07/2026
	OpenAI	Via OpenRouteropenai/fast	$4	$24	Source09/07/2026
	OpenAI	Via OpenRouteropenai/flex	Lowest listed price: $1	Lowest listed price: $6	Source09/07/2026
GPT-5.5	OpenAI	Direct from developerUp to 272,000 context tokens	$5	$30	Source07/30/2026
	OpenAI	Direct from developerAbove 272,000 context tokens	$10	$45	Source07/30/2026
	Amazon Bedrock	Via OpenRouteramazon-bedrock/us-east-1	$5.5	$33	Source09/07/2026
	Azure	Via OpenRouterazure	$5	$30	Source09/07/2026
	Azure	Via OpenRouterazure/eu	$5.5	$33	Source09/07/2026
	Azure	Via OpenRouterazure/us	$5.5	$33	Source09/07/2026
	OpenAI	Via OpenRouteropenai	$5	$30	Source09/07/2026
	OpenAI	Via OpenRouteropenai/fast	$12.5	$75	Source09/07/2026
	OpenAI	Via OpenRouteropenai/flex	Lowest listed price: $2.5	Lowest listed price: $15	Source09/07/2026

Performance

Shared benchmarks

Only published results with the same benchmark version, task, metric, and comparison cohort are paired. Different reasoning levels, agents, harnesses, or output limits appear directly in the table.

Capability profile from matched benchmarks

Each axis averages directly matched benchmark families. A value of 100 means the stronger value within this pair, not a universal quality score.

GPT-5.6 Terra

GPT-5.5

Coding: 93.2 / 96.8 (5 benchmark families)

Reasoning: 97.2 / 99.8 (2 benchmark families)

Knowledge: 98.2 / 99.9 (1 benchmark family)

Finance: 98.1 / 97.9 (5 benchmark families)

Legal: 95.8 / 98.5 (3 benchmark families)

Agentic tasks: 96.8 / 99.3 (2 benchmark families)

Multimodal: 98 / 100 (1 benchmark family)

Search: 100 / 96.5 (1 benchmark family)

171 matched rows from 26 benchmark families were considered. The radar shows 8 of 12 comparable categories.

GPT-5.6 TerraGPT-5.5

BrowseCompOverall. Higher is better. No documented setup difference

GPT-5.6 Terra87.5 %

GPT-5.584.4 %

Code MigrationOverall. Higher is better. No documented setup difference

GPT-5.6 Terra47.804 %

GPT-5.545.161 %

CorpFin v2Overall. Higher is better. No documented setup difference

GPT-5.6 Terra65.307 %

GPT-5.568.415 %

EMBOverall. Higher is better. No documented setup difference

GPT-5.6 Terra66.205 %

GPT-5.564.543 %

Finance Agent (v2)Overall. Higher is better. No documented setup difference

GPT-5.6 Terra54.436 %

GPT-5.551.76 %

GPQA DiamondOverall. Higher is better. No documented setup difference

GPT-5.6 Terra90.909 %

GPT-5.593.182 %

Harvey's Legal Agent BenchmarkOverall · Task fully resolved. Higher is better. No documented setup difference

GPT-5.6 Terra0.833 %

GPT-5.53.75 %

Legal Research BenchOverall · All-pass. Higher is better. No documented setup difference

GPT-5.6 Terra41.346 %

GPT-5.540.385 %

LegalBenchOverall. Higher is better. No documented setup difference

GPT-5.6 Terra85.105 %

GPT-5.586.515 %

LiveBenchOverall. Higher is better. No documented setup difference

GPT-5.6 Terra77.94 points

GPT-5.580.19 points

LiveCodeBench v6Overall. Higher is better. No documented setup difference

GPT-5.6 Terra85.93 %

GPT-5.585.296 %

LMArena Text, Style ControlOverall. Higher is better. Test setup not documented as identical

GPT-5.6 Terra1,466.233

GPT-5.51,477.235

Key to the marks:better valuebetter, but not conclusivetie

Benchmark scores for GPT-5.6 Terra and GPT-5.5
Benchmark and task	GPT-5.6 Terra	GPT-5.5	Source
BrowseCompOverall Test details No documented difference in the test setupVersion: 1,266 tasksMetric: accuracyScoring: Higher is betterCross-vendor comparison table with browsing tools. Exact agent systems differ.	Winner87.5 % Editorial selection from the vendor table, not a complete extract of the comparison cohort.	84.4 % Editorial selection from the vendor table, not a complete extract of the comparison cohort.	OpenAI Source details Data as of: 07/09/2026
Code MigrationOverall Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is better	Winner47.804 %	45.161 %	Vals AI Source details Data as of: 09/03/2026
CorpFin v2Overall Test details No documented difference in the test setupVersion: 2Metric: accuracyScoring: Higher is better	65.307 %	Winner68.415 %	Vals AI Source details Data as of: 08/12/2026
EMBOverall Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is better	Winner66.205 %	64.543 %	Vals AI Source details Data as of: 09/03/2026
Finance Agent (v2)Overall Test details No documented difference in the test setupVersion: 2Metric: accuracyScoring: Higher is better	Winner54.436 %	51.76 %	Vals AI Source details Data as of: 09/03/2026
GPQA DiamondOverall Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is better	90.909 %	Winner93.182 %	Vals AI Source details Data as of: 09/01/2026
Harvey's Legal Agent BenchmarkOverall · Task fully resolved Test details No documented difference in the test setupVersion: 1Metric: task resolution rateScoring: Higher is better	0.833 %	Winner3.75 %	Vals AI Source details Data as of: 09/03/2026
Legal Research BenchOverall · All-pass Test details No documented difference in the test setupVersion: 1Metric: all-pass rateScoring: Higher is better	Winner41.346 %	40.385 %	Vals AI Source details Data as of: 09/03/2026
LegalBenchOverall Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is better	85.105 %	Winner86.515 %	Vals AI Source details Data as of: 09/01/2026
LiveBenchOverall Test details No documented difference in the test setupVersion: 2026-06-25Metric: composite scoreScoring: Higher is better23 objectively scored tasks across seven categories.	77.94 points	Winner80.19 points	LiveBench Source details Data as of: 06/25/2026
LiveCodeBench v6Overall Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is better	Winner85.93 %	85.296 %	Vals AI Source details Data as of: 09/01/2026
LMArena Text, Style ControlOverall Test details Test setup not documented as identicalNo direct winnerVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterGPT-5.6 Terra: Snapshot: gpt-5.6-terra-xhigh, Reasoning: xhigh, Harness: Arena text, style control GPT-5.5: Snapshot: gpt-5.5, Harness: Arena text, style controlGPT-5.6 Terra: Style-controlled Arena rating from 23,942 anonymous pairwise votes.GPT-5.5: Style-controlled Arena rating from 64,686 anonymous pairwise votes.	1,466.233 95% confidence interval: ±5.062The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	Higher measured value1,477.235 95% confidence interval: ±3.839The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	LMArena Source details Data as of: 09/01/2026
MedCodeOverall Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is better	43.414 %	Winner49.1 %	Vals AI Source details Data as of: 09/03/2026
MedScribeOverall Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is better	82.869 %	Winner86.868 %	Vals AI Source details Data as of: 09/03/2026
MMLU-ProOverall Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is better	86.659 %	Winner88.144 %	Vals AI Source details Data as of: 09/01/2026
MMMU-ProOverall Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is better	86.474 %	Winner88.266 %	Vals AI Source details Data as of: 09/01/2026
MortgageTaxOverall Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is better	67.329 %	Winner68.76 %	Vals AI Source details Data as of: 09/01/2026
ProgramBenchOverall Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is better	Tie0.5 %	Tie0.5 %	Vals AI Source details Data as of: 09/01/2026
Public Benefits Bench v1.1Overall Test details No documented difference in the test setupVersion: 1.1Metric: accuracyScoring: Higher is better	Winner62.382 %	60.893 %	Vals AI Source details Data as of: 09/01/2026
SAGEOverall Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is better	47.003 %	Winner51.532 %	Vals AI Source details Data as of: 09/03/2026
SkillsBenchOverall Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is betterSettings: Harness: OpenHands	58.895 %	Winner62.209 %	Vals AI Source details Data as of: 09/01/2026
SWE-bench VerifiedOverall Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is better	Winner95.4 %	82.6 %	Vals AI Source details Data as of: 09/01/2026
TaxEval v2Overall Test details No documented difference in the test setupVersion: 2Metric: accuracyScoring: Higher is better	Winner76.166 %	74.98 %	Vals AI Source details Data as of: 09/01/2026
Terminal-Bench 2.1Overall Test details No documented difference in the test setupVersion: 2.1Metric: accuracyScoring: Higher is better	Winner77.528 %	76.404 %	Vals AI Source details Data as of: 09/03/2026
Vals IndexOverall Test details No documented difference in the test setupVersion: 2Metric: weighted index scoreScoring: Higher is better	Winner59.588 %	57.411 %	Vals AI Source details Data as of: 09/03/2026
Vals Multimodal IndexOverall Test details No documented difference in the test setupVersion: 1.2Metric: weighted index scoreScoring: Higher is better	65.07 %	Winner68.071 %	Vals AI Source details Data as of: 08/11/2026
Vibe Code Bench v1.1Overall Test details No documented difference in the test setupVersion: 1.1Metric: accuracyScoring: Higher is betterSettings: Harness: OpenHands	Winner74.594 %	69.847 %	Vals AI Source details Data as of: 09/03/2026
Code MigrationCLI Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is betterSettings: Reasoning: xhigh, Output limit: 128,000 tokens	26.645 %	Winner36.882 %	Vals AI Source details Data as of: 08/09/2026
CorpFin v2Exact Pages Test details No documented difference in the test setupVersion: 2Metric: accuracyScoring: Higher is betterSettings: Reasoning: xhigh, Output limit: 128,000 tokens	63.17 %	Winner65.618 %	Vals AI Source details Data as of: 08/06/2026
EMBComps Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is betterSettings: Reasoning: xhigh, Output limit: 128,000 tokens	Winner72.539 %	64.681 %	Vals AI Source details Data as of: 08/06/2026
Finance Agent (v2)Adjustments Test details No documented difference in the test setupVersion: 2Metric: accuracyScoring: Higher is betterSettings: Reasoning: xhigh, Output limit: 128,000 tokens	Winner50.442 %	45.74 %	Vals AI Source details Data as of: 08/06/2026
GPQA DiamondFew-Shot CoT Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is betterSettings: Reasoning: xhigh, Output limit: 128,000 tokens	89.899 %	Winner93.434 %	Vals AI Source details Data as of: 08/09/2026
Harvey's Legal Agent BenchmarkAntitrust Competition · Criteria passed Test details No documented difference in the test setupVersion: 1Metric: criteria pass rateScoring: Higher is betterSettings: Reasoning: xhigh, Output limit: 128,000 tokens	76.449 %	Winner82.246 %	Vals AI Source details Data as of: 08/07/2026
Legal Research BenchAdministrative / Regulatory · All-pass Test details No documented difference in the test setupVersion: 1Metric: all-pass rateScoring: Higher is betterSettings: Reasoning: xhigh, Output limit: 128,000 tokens	Winner57.576 %	53.03 %	Vals AI Source details Data as of: 08/06/2026
LegalBenchConclusion Tasks Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is betterSettings: Reasoning: xhigh, Output limit: 128,000 tokens	Winner90.327 %	89.644 %	Vals AI Source details Data as of: 08/09/2026
LiveBenchAgentic coding Test details No documented difference in the test setupVersion: 2026-06-25Metric: category scoreScoring: Higher is betterAgentic coding category inside the same LiveBench release.	Winner54.9 points	54 points	LiveBench Source details Data as of: 06/25/2026
LiveCodeBench v6Easy Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is betterSettings: Reasoning: xhigh, Output limit: 128,000 tokens	97.205 %	Winner98.447 %	Vals AI Source details Data as of: 08/09/2026
LMArena Text, Style ControlOverall Test details Test setup not documented as identicalNo direct winnerVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterGPT-5.6 Terra: Snapshot: gpt-5.6-terra-xhigh, Reasoning: xhigh, Harness: Arena text, style control GPT-5.5: Snapshot: gpt-5.5-high, Reasoning: high, Harness: Arena text, style controlGPT-5.6 Terra: Style-controlled Arena rating from 23,942 anonymous pairwise votes.GPT-5.5: Style-controlled Arena rating from 63,265 anonymous pairwise votes.	1,466.233 95% confidence interval: ±5.062The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	Higher measured value1,482.334 95% confidence interval: ±3.856The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	LMArena Source details Data as of: 09/01/2026
MMLU-ProBiology Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is betterSettings: Reasoning: xhigh, Output limit: 128,000 tokens	93.026 %	Winner94.003 %	Vals AI Source details Data as of: 08/09/2026
MortgageTaxNumerical Extraction Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is betterSettings: Reasoning: xhigh, Output limit: 128,000 tokens	74.324 %	Winner76.153 %	Vals AI Source details Data as of: 08/09/2026
ProgramBenchAlmost Resolved Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is betterSettings: Reasoning: xhigh, Output limit: 128,000 tokens	6.5 %	Winner9.5 %	Vals AI Source details Data as of: 08/09/2026
SAGEAutomata Test details Test setup not documented as identicalNo direct winnerVersion: 1Metric: accuracyScoring: Higher is betterGPT-5.6 Terra: Reasoning: xhigh, Output limit: 30,000 tokens GPT-5.5: Reasoning: xhigh, Temperature: 1, Output limit: 30,000 tokens	38.889 %	Higher measured value40.476 %	Vals AI Source details Data as of: 08/09/2026
SkillsBenchNo Skills Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is betterSettings: Reasoning: xhigh, Harness: OpenHands, Output limit: 128,000 tokens	44.208 %	Winner49.875 %	Vals AI Source details Data as of: 08/07/2026
SWE-bench Verified<15 min fix Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is betterSettings: Reasoning: xhigh, Output limit: 128,000 tokens	77.835 %	Winner91.753 %	Vals AI Source details Data as of: 08/08/2026
TaxEval v2Correctness Test details No documented difference in the test setupVersion: 2Metric: accuracyScoring: Higher is betterSettings: Reasoning: xhigh, Output limit: 128,000 tokens	Winner65.822 %	65.495 %	Vals AI Source details Data as of: 08/09/2026
Terminal-Bench 2.1Easy Test details Test setup not documented as identicalNo direct winnerVersion: 2.1Metric: accuracyScoring: Higher is betterGPT-5.6 Terra: Reasoning: xhigh, Output limit: 128,000 tokens GPT-5.5: Reasoning: high, Output limit: 128,000 tokens	No direct winner83.333 %	No direct winner83.333 %	Vals AI Source details Data as of: 08/06/2026
Vals IndexCorpFin v2 Test details No documented difference in the test setupVersion: 1.2Metric: benchmark scoreScoring: Higher is betterSettings: Reasoning: xhigh, Output limit: 128,000 tokens	66.434 %	Winner69.58 %	Vals AI Source details Data as of: 08/09/2026
Vals Multimodal IndexCorpFin v2 Test details No documented difference in the test setupVersion: 1.2Metric: benchmark scoreScoring: Higher is betterSettings: Reasoning: xhigh, Output limit: 128,000 tokens	66.434 %	Winner69.58 %	Vals AI Source details Data as of: 08/09/2026
Code MigrationCOBOL Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is betterSettings: Reasoning: xhigh, Output limit: 128,000 tokens	65.844 %	Winner70 %	Vals AI Source details Data as of: 08/09/2026
CorpFin v2Max Fitting Context Test details No documented difference in the test setupVersion: 2Metric: accuracyScoring: Higher is betterSettings: Reasoning: xhigh, Output limit: 128,000 tokens	66.317 %	Winner70.047 %	Vals AI Source details Data as of: 08/06/2026
EMBDataroom Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is betterSettings: Reasoning: xhigh, Output limit: 128,000 tokens	Winner81.591 %	81.227 %	Vals AI Source details Data as of: 08/06/2026
Finance Agent (v2)All-Pass Test details No documented difference in the test setupVersion: 2Metric: accuracyScoring: Higher is betterSettings: Reasoning: xhigh, Output limit: 128,000 tokens	Winner40.448 %	39.562 %	Vals AI Source details Data as of: 08/06/2026
GPQA DiamondZero-Shot CoT Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is betterSettings: Reasoning: xhigh, Output limit: 128,000 tokens	91.919 %	Winner92.929 %	Vals AI Source details Data as of: 08/09/2026
Harvey's Legal Agent BenchmarkAntitrust Competition · Task fully resolved Test details No documented difference in the test setupVersion: 1Metric: task resolution rateScoring: Higher is betterSettings: Reasoning: xhigh, Output limit: 128,000 tokens	Tie0 %	Tie0 %	Vals AI Source details Data as of: 08/07/2026
Legal Research BenchAdministrative / Regulatory · Weighted partial credit Test details No documented difference in the test setupVersion: 1Metric: weighted partial-credit scoreScoring: Higher is betterSettings: Reasoning: xhigh, Output limit: 128,000 tokens	Winner86.696 %	86.669 %	Vals AI Source details Data as of: 08/06/2026
LegalBenchInterpretation Tasks Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is betterSettings: Reasoning: xhigh, Output limit: 128,000 tokens	83.178 %	Winner83.836 %	Vals AI Source details Data as of: 08/09/2026
LiveBenchCoding Test details No documented difference in the test setupVersion: 2026-06-25Metric: category scoreScoring: Higher is betterCoding category inside the same LiveBench release.	78.2 points	Winner82.1 points	LiveBench Source details Data as of: 06/25/2026
LiveCodeBench v6Hard Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is betterSettings: Reasoning: xhigh, Output limit: 128,000 tokens	Winner72.857 %	69.714 %	Vals AI Source details Data as of: 08/09/2026
LMArena Text, Style ControlBusiness, management and finance Test details Test setup not documented as identicalNo direct winnerVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterGPT-5.6 Terra: Snapshot: gpt-5.6-terra-xhigh, Reasoning: xhigh, Harness: Arena text, style control GPT-5.5: Snapshot: gpt-5.5, Harness: Arena text, style controlGPT-5.6 Terra: Style-controlled Arena rating from 4,559 anonymous pairwise votes.GPT-5.5: Style-controlled Arena rating from 12,904 anonymous pairwise votes.	1,467.133 95% confidence interval: ±9.403The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	Higher measured value1,485.915 95% confidence interval: ±6.577The Arena score measures human preference rather than a fixed capability test. The confidence interval and model configuration are part of the result.	LMArena Source details Data as of: 09/01/2026
MMLU-ProBusiness Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is betterSettings: Reasoning: xhigh, Output limit: 128,000 tokens	90.875 %	Winner92.015 %	Vals AI Source details Data as of: 08/09/2026
MortgageTaxSemantic Extraction Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is betterSettings: Reasoning: xhigh, Output limit: 128,000 tokens	60.334 %	Winner61.367 %	Vals AI Source details Data as of: 08/09/2026
ProgramBenchFully Resolved Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is betterSettings: Reasoning: xhigh, Output limit: 128,000 tokens	Tie0.5 %	Tie0.5 %	Vals AI Source details Data as of: 08/09/2026
SAGECalculus Test details Test setup not documented as identicalNo direct winnerVersion: 1Metric: accuracyScoring: Higher is betterGPT-5.6 Terra: Reasoning: xhigh, Output limit: 30,000 tokens GPT-5.5: Reasoning: xhigh, Temperature: 1	59.259 %	Higher measured value72.222 %	Vals AI Source details Data as of: 08/09/2026
SkillsBenchWith Skills Test details No documented difference in the test setupVersion: 1Metric: accuracyScoring: Higher is betterSettings: Reasoning: xhigh, Harness: OpenHands, Output limit: 128,000 tokens	60.645 %	Winner62.209 %	Vals AI Source details Data as of: 08/07/2026
Key to the marks:better valuebetter, but not conclusivetie

How to read this comparison

Winning one benchmark is not an overall verdict

Your workload, budget, and required context length matter most. A coding benchmark says little about visual understanding or agent performance.

The standard price table uses direct developer rates. The provider comparison labels OpenRouter endpoints, regions, and special tariffs separately.

Every result links to its measurement source and retrieval date. Labels distinguish vendor reports, official benchmark leaderboards, and independent evaluations. Disputed or archived results are excluded.

More matchups

Compare GPT-5.6 Terra and GPT-5.5 with other leading models using the same data and benchmark logic.

View the complete LLM comparison