Models · OpenAI · Out sinceReleased 22 Sep 2026

GPT-6 Sol#10 for coding.#10 for coding, best at max effort.

Available on OpenRouterOn OpenRouterAvailable nowGenerally availableClosed (can't be downloaded)Proprietary

Where you can use it:In apps:ChatGPT (Business plan) · GPT-6 Sol (Work/Codex only)

GPT-6 Sol is made by OpenAI. Among the models we track it ranks #10 for coding, #28 for writing, #31 for research and analysis. It's mid-priced to use.

Superseded by GPT-6.1 Sol a week later. Default effort medium. Cached input $0.20/M. Knowledge cutoff Apr 20, 2026. System card link on launch post points to the GPT-6 Astra system card.

Writing & creativity: 36.5 / 100 · #28
Research & analysis: 49.7 / 100 · #31
Coding: 64.0 / 100 · #10
Price: Mid-priced$2 / $10
Price per 1M (blended)Blended / 1M: $4
MemoryContext: 1.1M
Longest answerMax output: 128K
Test resultsResults: 138 (103 independent103 indep.)
Out sinceReleased: 22 Sep 2026
Made byVendor: OpenAI
UnderstandsInputs: text, image

Scores are out of 100 for each area, compared with every model we track; “#” is its rank. Context is how much text it can read at once (1M is roughly 700,000 words). Blended price mixes the cost of what you send and what it writes back.

Thinking levels

Reasoning settings

No reasoningLowMediumHighExtra highMax

Highlighted: where it did best for coding (Maximum thinking).

Highlighted: dominant setting in its coding composite.

Links

Using it through OpenRouter

On OpenRouter

OpenRouter is a service that gives access to many AI models in one place. This model is listed there as openai/gpt-6-sol.

Route it as openai/gpt-6-sol at $2 in / $10 out per 1M tokens, 1.1M context. Listed since 22 Sep 2026.

include_reasoningmax_completion_tokensmax_tokensreasoningreasoning_effortresponse_formatseedstructured_outputstool_choicetools

Thinking level

Reasoning effort

How long should GPT-6 SolWhere GPT-6 Solthink?peaks.

Each chart is one test. Left to right, the model thinks longer; higher is a better result. The yellow ring marks where it did best. Thinking level is how long the AI thinks before it answers. Higher is slower and costs more, and it isn't always better.

Score per reasoning setting on each benchmark with ≥2 settings; ring = peak. Canonical results (independent over vendor).

Terminal-Bench 4.0

Best at MaxBest at Max

DeepSWE

Best at MaxBest at Max

FrontierCode

Best at MaxBest at Max

SciCode

Best at MaxBest at Max

Agents' Last Exam

Best at MaxBest at Max

Artificial Analysis Intelligence Index

Best at MaxBest at Max

Terminal-Bench-Science 0.1

Best at MaxBest at Max

AutomationBench

Best at Extra high, worse abovePeaks at Extra high

OSWorld 2.0 offline set

Best at MaxBest at Max

AA-Briefcase v1.1

Best at MaxBest at Max

AA-LCR

Best at HighBest at High

AA-Omniscience Accuracy

Best at MaxBest at Max

AA-Omniscience Hallucination Rate

Best at Low, worse abovePeaks at Low

Lower is better on this test

AA-Omniscience Index

Best at MaxBest at Max

ARC-AGI-3

Best at MaxBest at Max

GDP.pdf

Best at High, worse abovePeaks at High

Humanity's Last Exam

Best at MaxBest at Max

Every test result

Every result

All the numbers,with where they came from.

Every result we've found for this model, with who measured it and a link to where we read it.

All raw rows (vendor and independent kept separate), with setting label, source, date, harness and notes.

All results for this model
						NotesNotes
AA-Briefcase v1.1	905	Lowreasoning effort=low	Independent testIndependentArtificial Analysis (via Anthropic launch post) ↗	28 Sep 2026	—	AA-Briefcase v1.1 (Artificial Analysis, long-horizon knowledge work, Elo); cost/task $0.12
AA-Briefcase v1.1	1142	Mediumreasoning effort=med	Independent testIndependentArtificial Analysis (via Anthropic launch post) ↗	28 Sep 2026	—	AA-Briefcase v1.1 (Artificial Analysis, long-horizon knowledge work, Elo); cost/task $0.34
AA-Briefcase v1.1	1289	Highreasoning effort=high	Independent testIndependentArtificial Analysis (via Anthropic launch post) ↗	28 Sep 2026	—	AA-Briefcase v1.1 (Artificial Analysis, long-horizon knowledge work, Elo); cost/task $0.63
AA-Briefcase v1.1	1364	Extra highreasoning effort=xhigh	Independent testIndependentArtificial Analysis (via Anthropic launch post) ↗	28 Sep 2026	—	AA-Briefcase v1.1 (Artificial Analysis, long-horizon knowledge work, Elo); cost/task $1.19
AA-Briefcase v1.1	1483	Maxreasoning effort=max	Independent testIndependentArtificial Analysis (via Anthropic launch post) ↗	28 Sep 2026	—	AA-Briefcase v1.1 (Artificial Analysis, long-horizon knowledge work, Elo); cost/task $2.67
AA-Briefcase v1.1	1479	MaxGPT-6 Sol (Max)	Independent testIndependentArtificial Analysis ↗	1 Oct 2026	—
AA-LCR	79.3%	LowGPT-6 Sol (Low)	Independent testIndependentArtificial Analysis ↗	1 Oct 2026	—	AA-LCR accuracy, AA-run (long multi-document reasoning)
AA-LCR	82.3%	MediumGPT-6 Sol (Medium)	Independent testIndependentArtificial Analysis ↗	1 Oct 2026	—	AA-LCR accuracy, AA-run (long multi-document reasoning)
AA-LCR	83.7%	HighGPT-6 Sol (High)	Independent testIndependentArtificial Analysis ↗	1 Oct 2026	—	AA-LCR accuracy, AA-run (long multi-document reasoning)
AA-LCR	81.3%	Extra highGPT-6 Sol (Xhigh)	Independent testIndependentArtificial Analysis ↗	1 Oct 2026	—	AA-LCR accuracy, AA-run (long multi-document reasoning)
AA-LCR	83.7%	MaxGPT-6 Sol (Max)	Independent testIndependentArtificial Analysis ↗	1 Oct 2026	—	AA-LCR accuracy, AA-run (long multi-document reasoning)
AA-Omniscience Accuracy	51.3%	LowGPT-6 Sol (Low)	Independent testIndependentArtificial Analysis ↗	1 Oct 2026	—
AA-Omniscience Accuracy	53.5%	MediumGPT-6 Sol (Medium)	Independent testIndependentArtificial Analysis ↗	1 Oct 2026	—
AA-Omniscience Accuracy	53.7%	HighGPT-6 Sol (High)	Independent testIndependentArtificial Analysis ↗	1 Oct 2026	—
AA-Omniscience Accuracy	53.9%	Extra highGPT-6 Sol (Xhigh)	Independent testIndependentArtificial Analysis ↗	1 Oct 2026	—
AA-Omniscience Accuracy	54.5%	MaxGPT-6 Sol (Max)	Independent testIndependentArtificial Analysis ↗	1 Oct 2026	—
AA-Omniscience Hallucination Rate	50.7%	LowGPT-6 Sol (Low)	Independent testIndependentArtificial Analysis ↗	1 Oct 2026	—	share of non-correct answers that were wrong instead of abstaining; = 1 - AA 'omniscienceNonHallucination' field (equals breakdown.hallucinationRate on the eval page)
AA-Omniscience Hallucination Rate	56.8%	MediumGPT-6 Sol (Medium)	Independent testIndependentArtificial Analysis ↗	1 Oct 2026	—	share of non-correct answers that were wrong instead of abstaining; = 1 - AA 'omniscienceNonHallucination' field (equals breakdown.hallucinationRate on the eval page)
AA-Omniscience Hallucination Rate	58.1%	HighGPT-6 Sol (High)	Independent testIndependentArtificial Analysis ↗	1 Oct 2026	—	share of non-correct answers that were wrong instead of abstaining; = 1 - AA 'omniscienceNonHallucination' field (equals breakdown.hallucinationRate on the eval page)
AA-Omniscience Hallucination Rate	58.9%	Extra highGPT-6 Sol (Xhigh)	Independent testIndependentArtificial Analysis ↗	1 Oct 2026	—	share of non-correct answers that were wrong instead of abstaining; = 1 - AA 'omniscienceNonHallucination' field (equals breakdown.hallucinationRate on the eval page)
AA-Omniscience Hallucination Rate	60.1%	MaxGPT-6 Sol (Max)	Independent testIndependentArtificial Analysis ↗	1 Oct 2026	—	share of non-correct answers that were wrong instead of abstaining; = 1 - AA 'omniscienceNonHallucination' field (equals breakdown.hallucinationRate on the eval page)
AA-Omniscience Index	26.5	LowGPT-6 Sol (Low)	Independent testIndependentArtificial Analysis ↗	1 Oct 2026	—	AA-Omniscience Index (-100..100): correct minus incorrect, abstentions not penalised
AA-Omniscience Index	27	MediumGPT-6 Sol (Medium)	Independent testIndependentArtificial Analysis ↗	1 Oct 2026	—	AA-Omniscience Index (-100..100): correct minus incorrect, abstentions not penalised
AA-Omniscience Index	26.8	HighGPT-6 Sol (High)	Independent testIndependentArtificial Analysis ↗	1 Oct 2026	—	AA-Omniscience Index (-100..100): correct minus incorrect, abstentions not penalised
AA-Omniscience Index	26.7	Extra highGPT-6 Sol (Xhigh)	Independent testIndependentArtificial Analysis ↗	1 Oct 2026	—	AA-Omniscience Index (-100..100): correct minus incorrect, abstentions not penalised
AA-Omniscience Index	27.1	MaxGPT-6 Sol (Max)	Independent testIndependentArtificial Analysis ↗	1 Oct 2026	—	AA-Omniscience Index (-100..100): correct minus incorrect, abstentions not penalised
Agents' Last Exam	48.7%	Lowreasoning effort=low	Maker's own figureVendor-reportedOpenAI ↗	22 Sep 2026	—	Agents' Last Exam V1 (long-horizon professional workflows, 55 sub-industries); vendor-estimated API cost/task $0.8641
Agents' Last Exam	53.1%	Mediumreasoning effort=medium	Maker's own figureVendor-reportedOpenAI ↗	22 Sep 2026	—	Agents' Last Exam V1 (long-horizon professional workflows, 55 sub-industries); vendor-estimated API cost/task $1.2664
Agents' Last Exam	52.6%	Highreasoning effort=high	Maker's own figureVendor-reportedOpenAI ↗	22 Sep 2026	—	Agents' Last Exam V1 (long-horizon professional workflows, 55 sub-industries); vendor-estimated API cost/task $1.5302
Agents' Last Exam	55.4%	Extra highreasoning effort=xhigh	Maker's own figureVendor-reportedOpenAI ↗	22 Sep 2026	—	Agents' Last Exam V1 (long-horizon professional workflows, 55 sub-industries); vendor-estimated API cost/task $1.668
Agents' Last Exam	56.4%	Maxreasoning effort=max	Maker's own figureVendor-reportedOpenAI ↗	22 Sep 2026	—	Agents' Last Exam V1 (long-horizon professional workflows, 55 sub-industries); vendor-estimated API cost/task $2.9313
ARC-AGI-2	89.6%	MaxGPT-6 Sol (Max)	Independent testIndependentARC Prize ↗	30 Sep 2026	—	semi-private set; $0.44/task; data https://arcprize.org/media/data/leaderboard/v2.json
ARC-AGI-3	1.2%	No reasoningGPT-6 Sol - Provider Adapter (None)	Independent testIndependentARC Prize ↗	30 Sep 2026	Provider Adapter (vendor-built agent harness)	semi-private set; total cost $4,234; data https://arcprize.org/media/data/leaderboard/v3.json
ARC-AGI-3	0.3%	No reasoningGPT-6 Sol (None)	Independent testIndependentARC Prize ↗	30 Sep 2026	—	semi-private set; total cost $4,656; data https://arcprize.org/media/data/leaderboard/v3.json
ARC-AGI-3	1.8%	LowGPT-6 Sol - Provider Adapter (Low)	Independent testIndependentARC Prize ↗	30 Sep 2026	Provider Adapter (vendor-built agent harness)	semi-private set; total cost $4,745; data https://arcprize.org/media/data/leaderboard/v3.json
ARC-AGI-3	4.9%	MediumGPT-6 Sol - Provider Adapter (Medium)	Independent testIndependentARC Prize ↗	30 Sep 2026	Provider Adapter (vendor-built agent harness)	semi-private set; total cost $5,889; data https://arcprize.org/media/data/leaderboard/v3.json
ARC-AGI-3	0.4%	MediumGPT-6 Sol (Medium)	Independent testIndependentARC Prize ↗	30 Sep 2026	—	semi-private set; total cost $4,864; data https://arcprize.org/media/data/leaderboard/v3.json
ARC-AGI-3	5.9%	HighGPT-6 Sol - Provider Adapter (High)	Independent testIndependentARC Prize ↗	30 Sep 2026	Provider Adapter (vendor-built agent harness)	semi-private set; total cost $5,908; data https://arcprize.org/media/data/leaderboard/v3.json
ARC-AGI-3	0.9%	HighGPT-6 Sol (High)	Independent testIndependentARC Prize ↗	30 Sep 2026	—	semi-private set; total cost $5,072; data https://arcprize.org/media/data/leaderboard/v3.json
ARC-AGI-3	9.5%	Extra highGPT-6 Sol - Provider Adapter (XHigh)	Independent testIndependentARC Prize ↗	30 Sep 2026	Provider Adapter (vendor-built agent harness)	semi-private set; total cost $6,907; data https://arcprize.org/media/data/leaderboard/v3.json
ARC-AGI-3	1.8%	Extra highGPT-6 Sol (XHigh)	Independent testIndependentARC Prize ↗	30 Sep 2026	—	semi-private set; total cost $5,106; data https://arcprize.org/media/data/leaderboard/v3.json
ARC-AGI-3	23.0%	MaxGPT-6 Sol - Provider Adapter (Max)	Independent testIndependentARC Prize ↗	30 Sep 2026	Provider Adapter (vendor-built agent harness)	semi-private set; total cost $8,722; data https://arcprize.org/media/data/leaderboard/v3.json
ARC-AGI-3	4.6%	MaxGPT-6 Sol (Max)	Independent testIndependentARC Prize ↗	30 Sep 2026	—	semi-private set; total cost $5,554; data https://arcprize.org/media/data/leaderboard/v3.json
Artificial Analysis Coding Agent Index	56.7	MaxGPT-6 Sol (max)	Independent testIndependentArtificial Analysis ↗	1 Oct 2026	Codex	agent Codex; components: DeepSWE v1.1 69.0, SWE-Atlas-QnA 57.5, Terminal-Bench v4 43.4; avg cost $2.99/task; avg wall time 22 min/task
Artificial Analysis Intelligence Index	34.2	LowGPT-6 Sol (Low)	Independent testIndependentArtificial Analysis ↗	1 Oct 2026	—	AA slug gpt-6-sol-low; list price $2/10 per 1M in/out; cost to run AA Intelligence Index $0.13/task
Artificial Analysis Intelligence Index	39.8	MediumGPT-6 Sol (Medium)	Independent testIndependentArtificial Analysis ↗	1 Oct 2026	—	AA slug gpt-6-sol-medium; list price $2/10 per 1M in/out; cost to run AA Intelligence Index $0.25/task
Artificial Analysis Intelligence Index	42.4	HighGPT-6 Sol (High)	Independent testIndependentArtificial Analysis ↗	1 Oct 2026	—	AA slug gpt-6-sol-high; list price $2/10 per 1M in/out; cost to run AA Intelligence Index $0.37/task
Artificial Analysis Intelligence Index	44.2	Extra highGPT-6 Sol (Xhigh)	Independent testIndependentArtificial Analysis ↗	1 Oct 2026	—	AA slug gpt-6-sol-xhigh; list price $2/10 per 1M in/out; cost to run AA Intelligence Index $0.52/task
Artificial Analysis Intelligence Index	47.6	MaxGPT-6 Sol (Max)	Independent testIndependentArtificial Analysis ↗	1 Oct 2026	—	AA slug gpt-6-sol; list price $2/10 per 1M in/out; cost to run AA Intelligence Index $1.04/task
Artificial Analysis output speed	56 tok/s	LowGPT-6 Sol (Low)	Independent testIndependentArtificial Analysis ↗	1 Oct 2026	—	tokens/sec (median output speed, first-party API); TTFT 1.9s; list price $2/10 per 1M in/out
Artificial Analysis output speed	63 tok/s	HighGPT-6 Sol (High)	Independent testIndependentArtificial Analysis ↗	1 Oct 2026	—	tokens/sec (median output speed, first-party API); TTFT 19.2s; list price $2/10 per 1M in/out
Artificial Analysis output speed	68 tok/s	Extra highGPT-6 Sol (Xhigh)	Independent testIndependentArtificial Analysis ↗	1 Oct 2026	—	tokens/sec (median output speed, first-party API); TTFT 57.8s; list price $2/10 per 1M in/out
Artificial Analysis output speed	74 tok/s	MaxGPT-6 Sol (Max)	Independent testIndependentArtificial Analysis ↗	1 Oct 2026	—	tokens/sec (median output speed, first-party API); TTFT 177.8s; list price $2/10 per 1M in/out
AutomationBench	21.2%	Lowreasoning effort=low	Maker's own figureVendor-reportedOpenAI ↗	29 Sep 2026	—	AutomationBench 1.0.6 (Zapier); vendor-estimated API cost/task $0.1861
AutomationBench	26.9%	Mediumreasoning effort=medium	Maker's own figureVendor-reportedOpenAI ↗	29 Sep 2026	—	AutomationBench 1.0.6 (Zapier); vendor-estimated API cost/task $0.2098
AutomationBench	31.2%	Highreasoning effort=high	Maker's own figureVendor-reportedOpenAI ↗	29 Sep 2026	—	AutomationBench 1.0.6 (Zapier); vendor-estimated API cost/task $0.2368
AutomationBench	33.2%	Extra highreasoning effort=xhigh	Maker's own figureVendor-reportedOpenAI ↗	29 Sep 2026	—	AutomationBench 1.0.6 (Zapier); vendor-estimated API cost/task $0.2746
AutomationBench	32.0%	Maxreasoning effort=max	Maker's own figureVendor-reportedOpenAI ↗	29 Sep 2026	—	AutomationBench 1.0.6 (Zapier); vendor-estimated API cost/task $0.3406
Chartography (no tools)	53.6%	Defaultas listed by Anthropic (effort not stated)	Independent testIndependentSurge AI (via Anthropic launch post) ↗	28 Sep 2026	—	no tools
DeepSWE	37.2%	Lowreasoning effort=low	Maker's own figureVendor-reportedOpenAI ↗	29 Sep 2026	Codex	DeepSWE v1.1 (113 long-horizon SWE tasks); GPT models run in Codex; vendor-estimated API cost/task $0.1623
DeepSWE	56.6%	Mediumreasoning effort=medium	Maker's own figureVendor-reportedOpenAI ↗	29 Sep 2026	Codex	DeepSWE v1.1 (113 long-horizon SWE tasks); GPT models run in Codex; vendor-estimated API cost/task $0.3798
DeepSWE	65.3%	Highreasoning effort=high	Maker's own figureVendor-reportedOpenAI ↗	29 Sep 2026	Codex	DeepSWE v1.1 (113 long-horizon SWE tasks); GPT models run in Codex; vendor-estimated API cost/task $0.6404
DeepSWE	66.6%	Extra highreasoning effort=xhigh	Maker's own figureVendor-reportedOpenAI ↗	29 Sep 2026	Codex	DeepSWE v1.1 (113 long-horizon SWE tasks); GPT models run in Codex; vendor-estimated API cost/task $1.0033
DeepSWE	69.0%	MaxGPT-6 Sol (max)	Independent testIndependentArtificial Analysis ↗	1 Oct 2026	Codex	AA Coding Agent Index component; pass@1 avg of 3 attempts
DeepSWE	68.8%	Maxreasoning effort=max	Maker's own figureVendor-reportedOpenAI ↗	29 Sep 2026	Codex	DeepSWE v1.1 (113 long-horizon SWE tasks); GPT models run in Codex; vendor-estimated API cost/task $2.7439
Design Arena (all categories)	1292	Defaultgpt-6-sol	Independent testIndependentDesign Arena ↗	1 Oct 2026	—	Bradley-Terry Elo; ±7.8 SE; 2044 battles; win rate 52%
Design Arena (fullstack)	1237	Defaultgpt-6-sol	Independent testIndependentDesign Arena ↗	1 Oct 2026	—	Bradley-Terry Elo; ±10.8 SE; 1154 battles; win rate 52.2%
EQ-Bench Creative Writing v3 (Elo)	2125	Default*gpt-6-sol	Independent testIndependentEQ-Bench ↗	1 Oct 2026	—	leaderboard rank 4; rubric score 16.51/20; slop 8.97; avg length 6182 chars; Elo judged by Claude Sonnet 4.6; effort not stated by EQ-Bench; marked new (*)
FrontierCode	37.3%	Lowreasoning effort=low	Maker's own figureVendor-reportedOpenAI ↗	22 Sep 2026	Codex	FrontierCode 1.1 Main (Cognition); GPT models run in Codex CLI; vendor-estimated API cost/task $0.453
FrontierCode	45.9%	Mediumreasoning effort=medium	Maker's own figureVendor-reportedOpenAI ↗	22 Sep 2026	Codex	FrontierCode 1.1 Main (Cognition); GPT models run in Codex CLI; vendor-estimated API cost/task $0.7966
FrontierCode	47.7%	Highreasoning effort=high	Maker's own figureVendor-reportedOpenAI ↗	22 Sep 2026	Codex	FrontierCode 1.1 Main (Cognition); GPT models run in Codex CLI; vendor-estimated API cost/task $1.0793
FrontierCode	48.5%	Extra highreasoning effort=xhigh	Maker's own figureVendor-reportedOpenAI ↗	22 Sep 2026	Codex	FrontierCode 1.1 Main (Cognition); GPT models run in Codex CLI; vendor-estimated API cost/task $1.3748
FrontierCode	49.3%	Maxgpt-6-sol_max	Independent testIndependentFrontierCode (Cognition) via Epoch AI ↗	1 Oct 2026	codex	read from Epoch AI benchmark_data.zip (frontiercode_external.csv); original leaderboard https://cognition.com/frontiercode; harness codex; Mean@5
FrontierCode	49.3%	Maxreasoning effort=max	Maker's own figureVendor-reportedOpenAI ↗	22 Sep 2026	Codex	FrontierCode 1.1 Main (Cognition); GPT models run in Codex CLI; vendor-estimated API cost/task $2.137
FrontierMath (Tiers 1-3)	89.8%	Maxgpt-6-sol_max	Independent testIndependentEpoch AI ↗	22 Sep 2026	—	FrontierMath Tiers 1-3 (v2), Epoch-run; stderr 1.8pt; data https://epoch.ai/data/benchmark_data.zip (frontiermath_tiers_1_3_v2.csv)
FrontierMath Tier 4	90.0%	Maxgpt-6-sol_max	Independent testIndependentEpoch AI ↗	22 Sep 2026	—	FrontierMath Tier 4 (v2), Epoch-run; stderr 5.2pt; data https://epoch.ai/data/benchmark_data.zip (frontiermath_tier_4_v2.csv)
GDP.pdf	21.8%	Lowreasoning effort=low	Maker's own figureVendor-reportedOpenAI ↗	29 Sep 2026	—	GDP.pdf (Surge AI; professional questions over complex PDFs); vendor-estimated API cost/task $0.33
GDP.pdf	25.4%	Mediumreasoning effort=medium	Maker's own figureVendor-reportedOpenAI ↗	29 Sep 2026	—	GDP.pdf (Surge AI; professional questions over complex PDFs); vendor-estimated API cost/task $0.34
GDP.pdf	28.0%	Highreasoning effort=high	Maker's own figureVendor-reportedOpenAI ↗	29 Sep 2026	—	GDP.pdf (Surge AI; professional questions over complex PDFs); vendor-estimated API cost/task $0.35
GDP.pdf	23.8%	Extra highreasoning effort=xhigh	Maker's own figureVendor-reportedOpenAI ↗	29 Sep 2026	—	GDP.pdf (Surge AI; professional questions over complex PDFs); vendor-estimated API cost/task $0.37
GDP.pdf	24.8%	Maxreasoning effort=max	Maker's own figureVendor-reportedOpenAI ↗	29 Sep 2026	—	GDP.pdf (Surge AI; professional questions over complex PDFs); vendor-estimated API cost/task $0.43
GDPval-AA v2.1	1505	MaxGPT-6 Sol (Max)	Independent testIndependentArtificial Analysis ↗	1 Oct 2026	—	95% CI 1480.19-1530.25
GDPval-AA v2.1	1487	Defaultas listed by Anthropic (effort not stated)	Independent testIndependentArtificial Analysis (via Anthropic launch post) ↗	28 Sep 2026	—	GDPval-AA v2.1 (Artificial Analysis, Elo from blind pairwise comparisons, 44 occupations); may predate OpenAI image-understanding bug fix for GPT-6 Sol
GPQA Diamond	94.3%	Maxgpt-6-sol_max	Independent testIndependentEpoch AI ↗	22 Sep 2026	—	Epoch-run GPQA Diamond; stderr 1.5pt; data https://epoch.ai/data/benchmark_data.zip (gpqa_diamond.csv)
HLE Diamond	33.8%	Defaultgpt-6-sol	Independent testIndependentScale AI SEAL ↗	1 Oct 2026	—	rank 6 (Scale rank accounts for CI); ±2.9; entry added 2026-04-08
Humanity's Last Exam	34.9%	LowGPT-6 Sol (Low)	Independent testIndependentArtificial Analysis ↗	1 Oct 2026	—	AA-run evaluation (AA slug gpt-6-sol-low)
Humanity's Last Exam	41.0%	MediumGPT-6 Sol (Medium)	Independent testIndependentArtificial Analysis ↗	1 Oct 2026	—	AA-run evaluation (AA slug gpt-6-sol-medium)
Humanity's Last Exam	44.1%	HighGPT-6 Sol (High)	Independent testIndependentArtificial Analysis ↗	1 Oct 2026	—	AA-run evaluation (AA slug gpt-6-sol-high)
Humanity's Last Exam	46.3%	Extra highGPT-6 Sol (Xhigh)	Independent testIndependentArtificial Analysis ↗	1 Oct 2026	—	AA-run evaluation (AA slug gpt-6-sol-xhigh)
Humanity's Last Exam	47.9%	MaxGPT-6 Sol (Max)	Independent testIndependentArtificial Analysis ↗	1 Oct 2026	—	AA-run evaluation (AA slug gpt-6-sol)
IOI (Vals)	82.6%	Maxreasoning_effort=max	Independent testIndependentVals.ai ↗	29 Sep 2026	—	±9.24 stderr; $2.703/test
LMArena Code Arena (WebDev)	1689	Maxgpt-6-sol-max	Independent testIndependentLMArena ↗	30 Sep 2026	—	Code Arena \| WebDev overall (agentic web-dev, raw); rank 7 (CI rank 5-10); 95% CI 1678-1701; 3197 votes
LMArena Text - Coding category	1523	Maxgpt-6-sol-max	Independent testIndependentLMArena ↗	30 Sep 2026	—	text arena coding category, style control on; rank 28 (CI rank 7-73); 95% CI 1508-1537; 1799 votes
LMArena Text - Creative Writing	1438	Maxgpt-6-sol-max	Independent testIndependentLMArena ↗	30 Sep 2026	—	rank 57 (rank range 24-88); 95% CI 1422.1-1454.1; 1532 votes; style-controlled
LMArena Text - Expert	1498	Maxgpt-6-sol-max	Independent testIndependentLMArena ↗	30 Sep 2026	—	rank 43 (rank range 9-93); 95% CI 1477.9-1518.9; 818 votes; style-controlled
LMArena Text - Hard Prompts	1484	Maxgpt-6-sol-max	Independent testIndependentLMArena ↗	30 Sep 2026	—	rank 55 (rank range 31-80); 95% CI 1475.3-1493.6; 4639 votes; style-controlled
LMArena Text - Instruction Following	1453	Maxgpt-6-sol-max	Independent testIndependentLMArena ↗	30 Sep 2026	—	rank 57 (rank range 32-84); 95% CI 1440.8-1464.6; 2514 votes; style-controlled
LMArena Text - Longer Query	1466	Maxgpt-6-sol-max	Independent testIndependentLMArena ↗	30 Sep 2026	—	rank 61 (rank range 35-92); 95% CI 1455.4-1477.3; 3182 votes; style-controlled
LMArena Text - Multi-Turn	1469	Maxgpt-6-sol-max	Independent testIndependentLMArena ↗	30 Sep 2026	—	rank 61 (rank range 16-101); 95% CI 1450.8-1486.5; 1148 votes; style-controlled
LMArena Text - Non-English	1446	Maxgpt-6-sol-max	Independent testIndependentLMArena ↗	30 Sep 2026	—	rank 60 (rank range 41-86); 95% CI 1436.4-1454.6; 4644 votes; style-controlled
LMArena Text - Occupational: Business, Management & Finance	1448	Maxgpt-6-sol-max	Independent testIndependentLMArena ↗	30 Sep 2026	—	rank 77 (rank range 36-118); 95% CI 1432.1-1464.0; 1435 votes; style-controlled
LMArena Text - Occupational: Legal & Government	1458	Maxgpt-6-sol-max	Independent testIndependentLMArena ↗	30 Sep 2026	—	rank 80 (rank range 20-143); 95% CI 1434.2-1481.3; 649 votes; style-controlled
LMArena Text - Occupational: Writing, Literature & Language	1440	Maxgpt-6-sol-max	Independent testIndependentLMArena ↗	30 Sep 2026	—	rank 69 (rank range 31-98); 95% CI 1425.8-1453.4; 2023 votes; style-controlled
OSWorld 2.0 offline set	43.9%	Lowreasoning effort=low	Maker's own figureVendor-reportedOpenAI ↗	29 Sep 2026	—	OSWorld 2.0 offline set, partial reward, v2026.08.08 release; vendor-estimated API cost/task $1.0117
OSWorld 2.0 offline set	54.0%	Mediumreasoning effort=medium	Maker's own figureVendor-reportedOpenAI ↗	29 Sep 2026	—	OSWorld 2.0 offline set, partial reward, v2026.08.08 release; vendor-estimated API cost/task $1.3789
OSWorld 2.0 offline set	58.3%	Highreasoning effort=high	Maker's own figureVendor-reportedOpenAI ↗	29 Sep 2026	—	OSWorld 2.0 offline set, partial reward, v2026.08.08 release; vendor-estimated API cost/task $1.7096
OSWorld 2.0 offline set	60.5%	Extra highreasoning effort=xhigh	Maker's own figureVendor-reportedOpenAI ↗	29 Sep 2026	—	OSWorld 2.0 offline set, partial reward, v2026.08.08 release; vendor-estimated API cost/task $2.3002
OSWorld 2.0 offline set	64.4%	Maxreasoning effort=max	Maker's own figureVendor-reportedOpenAI ↗	29 Sep 2026	—	OSWorld 2.0 offline set, partial reward, v2026.08.08 release; vendor-estimated API cost/task $3.3695
PRBench Legal (Scale)	38.6%	Defaultgpt-6-sol	Independent testIndependentScale AI (SEAL) ↗	25 Sep 2026	—	Scale rank 28; ±1.58 CI
ProgramBench (fully resolved)	2.0%	Maxreasoning_effort=max	Independent testIndependentVals.ai ↗	29 Sep 2026	mini-SWE-agent	±0.992 stderr; $11.673/test; strict fully-resolved rate
SciCode	50.2%	LowGPT-6 Sol (Low)	Independent testIndependentArtificial Analysis ↗	1 Oct 2026	—	AA-run evaluation (AA slug gpt-6-sol-low)
SciCode	53.8%	MediumGPT-6 Sol (Medium)	Independent testIndependentArtificial Analysis ↗	1 Oct 2026	—	AA-run evaluation (AA slug gpt-6-sol-medium)
SciCode	54.9%	HighGPT-6 Sol (High)	Independent testIndependentArtificial Analysis ↗	1 Oct 2026	—	AA-run evaluation (AA slug gpt-6-sol-high)
SciCode	55.1%	Extra highGPT-6 Sol (Xhigh)	Independent testIndependentArtificial Analysis ↗	1 Oct 2026	—	AA-run evaluation (AA slug gpt-6-sol-xhigh)
SciCode	57.6%	MaxGPT-6 Sol (Max)	Independent testIndependentArtificial Analysis ↗	1 Oct 2026	—	AA-run evaluation (AA slug gpt-6-sol)
SimpleBench	73.1%	DefaultGPT-6 Sol	Independent testIndependentSimpleBench ↗	24 Sep 2026	—	AVG@5, temp 0.7; rank 15th; human baseline 83.7%; data https://simple-bench.com/static/js/leaderboard-data.js
SimpleQA Verified	60.7%	Maxgpt-6-sol_max	Independent testIndependentEpoch AI ↗	22 Sep 2026	—	Epoch-run (no tools); ±1.55 stderr
SWE Atlas - Codebase QnA	57.5%	MaxGPT-6 Sol (max)	Independent testIndependentArtificial Analysis ↗	1 Oct 2026	Codex	AA Coding Agent Index component; pass@1 avg of 3 attempts
Terminal-Bench 2.1	83.2%	Maxreasoning_effort=max	Independent testIndependentVals.ai ↗	27 Sep 2026	Terminus 2	±1.297 stderr; $0.373/test
Terminal-Bench 4.0	9.1%	LowGPT-6 Sol (Low)	Independent testIndependentArtificial Analysis ↗	1 Oct 2026	—	AA-run evaluation (AA slug gpt-6-sol-low)
Terminal-Bench 4.0	18.7%	MediumGPT-6 Sol (Medium)	Independent testIndependentArtificial Analysis ↗	1 Oct 2026	—	AA-run evaluation (AA slug gpt-6-sol-medium)
Terminal-Bench 4.0	26.3%	HighGPT-6 Sol (High)	Independent testIndependentArtificial Analysis ↗	1 Oct 2026	—	AA-run evaluation (AA slug gpt-6-sol-high)
Terminal-Bench 4.0	30.3%	Extra highGPT-6 Sol (Xhigh)	Independent testIndependentArtificial Analysis ↗	1 Oct 2026	—	AA-run evaluation (AA slug gpt-6-sol-xhigh)
Terminal-Bench 4.0	44.4%	Maxreasoning_effort=max	Independent testIndependentVals.ai ↗	29 Sep 2026	mini-SWE-agent	±3.535 stderr; $5.793/test
Terminal-Bench 4.0	43.9%	MaxGPT-6 Sol (Max)	Independent testIndependentArtificial Analysis ↗	1 Oct 2026	—	AA-run evaluation (AA slug gpt-6-sol)
Terminal-Bench 4.0	43.4%	MaxGPT-6 Sol (max)	Independent testIndependentArtificial Analysis ↗	1 Oct 2026	Codex	AA Coding Agent Index component run in the vendor agent harness; pass@1 avg of 3 attempts
Terminal-Bench-Science 0.1	9.2%	Lowreasoning effort=low	Maker's own figureVendor-reportedOpenAI ↗	29 Sep 2026	Codex	Terminal-Bench Science 0.1; vendor-estimated API cost/task $2.9986
Terminal-Bench-Science 0.1	14.5%	Mediumreasoning effort=medium	Maker's own figureVendor-reportedOpenAI ↗	29 Sep 2026	Codex	Terminal-Bench Science 0.1; vendor-estimated API cost/task $4.4068
Terminal-Bench-Science 0.1	14.6%	Highreasoning effort=high	Maker's own figureVendor-reportedOpenAI ↗	29 Sep 2026	Codex	Terminal-Bench Science 0.1; vendor-estimated API cost/task $4.6325
Terminal-Bench-Science 0.1	25.3%	Extra highreasoning effort=xhigh	Maker's own figureVendor-reportedOpenAI ↗	29 Sep 2026	Codex	Terminal-Bench Science 0.1; vendor-estimated API cost/task $6.7698
Terminal-Bench-Science 0.1	27.6%	Maxreasoning effort=max	Maker's own figureVendor-reportedOpenAI ↗	29 Sep 2026	Codex	Terminal-Bench Science 0.1; vendor-estimated API cost/task $12.1803
Vals Code Migration	57.2%	Maxreasoning_effort=max	Independent testIndependentVals.ai ↗	29 Sep 2026	—	±4.213 stderr; $15.677/test
Vals Index	57.5%	Maxreasoning_effort=max	Independent testIndependentVals.ai ↗	30 Sep 2026	—	±1.012 stderr; $7.581/test
Vals Legal Research Bench	28.8%	Maxreasoning_effort=max	Independent testIndependentVals.ai ↗	29 Sep 2026	—	vals id openai/gpt-6-sol; rank 42/72; ±3.149 stderr; $4.897717/test
Vals Public Benefits Bench v1.1	56.6%	Maxreasoning_effort=max	Independent testIndependentVals.ai ↗	29 Sep 2026	—	vals id openai/gpt-6-sol; rank 35/45; ±1.289 stderr; $2.491613/test
Vals Vibe Code Bench	87.8%	Maxreasoning_effort=max	Independent testIndependentVals.ai ↗	29 Sep 2026	OpenHands	±2.526 stderr; $26.356/test
Vectara Hallucination Leaderboard (HHEM)	6.5%	Defaultopenai/gpt-6-sol	Independent testIndependentVectara ↗	22 Sep 2026	—	factual consistency 93.5 %; answer rate 100.0 %; avg summary 71.4 words; HHEM-2.3 judge; effort not stated (API default)
Vending-Bench 2	$14,428	Default	Independent testIndependentAndon Labs ↗	1 Oct 2026	—	final money balance after simulated year, arithmetic mean across runs; ±$1,051; rank 2; only top 10 rendered server-side (57 more behind 'Show more')