GPT-5.6 Sol

OpenAILegacy

Input price: $4.00; Per 1 million tokens (USD)
Output price: $20.00; Per 1 million tokens (USD)
Effort levels: 6total; Default: Med
Benchmark results: 53total; Benchmark types: 17

Model information

API model ID: gpt-5.6-sol
release date: 2026-07-09
effort: None / Low / Med (default) / High / Xhigh / Max
Effort source: Based on reasoning effort on the OpenAI model page (developers.openai.com/api/docs/models/gpt-5.6-sol).
Claude Code: unavailable
Codex: unknown
Availability notes: alias gpt-5.6 → gpt-5.6-sol. The Codex models page lists only the GPT-6 series. Claude Code's official model list (code.claude.com/docs/en/model-config alias table) contains only Anthropic models, so claudeCode is false (connections via a gateway are not counted as official options).
Availability source: https://developers.openai.com/api/docs/models/gpt-5.6-sol

Price (USD per 1 million tokens)

Prices do not vary by effort; token consumption does.

input	$4.00
output	$20.00
cache read	$0.40
cache write	$5.00

Prompts with more than 272K input tokens are priced at 2x input and cache rates and 1.5x output for the full request. Batch/Flex 50%, Fast 2x. 'GPT-5.6 Sol's promotional pricing is available until November 30, 2026.' (Revised to $4/$20 on 2026-08-21)

| gpt-5.6-sol | $4.00 | $20.00 | $0.40 | $5.00 | $10.00 | $8.00 | $30.00 | $0.80 | $10.00 | $20.00 |

Source: https://developers.openai.com/api/docs/pricing

Benchmark results

Benchmark	effort	Score	Source cost (metric)	measured/reported by	source	Conditions and extraction method
Terminal-Bench 4.0Low	Low	7.9%	$1.46Cost per attempt	OpenAI	Introducing Claude Opus 5.5official · Retrieved 2026-11-12	Evidence (same value as comparison group Accuracy vs Cost figure announced by Anthropic (Opus 5.5 and Sonnet 5.5 pages))Vendor-reported value. A point in the figure "Terminal-Bench 4.0 / Accuracy vs Cost" on the Anthropic announcement page. values were taken from each point's aria-label (DOM); the page shows them only in the figure; value for a model from some other vendor. measured by: the figure makes no distinction / the table footnote says GPT series figures are reported values from OpenAI; the figure's effort label "Low" was mapped to low in the OpenAI API reasoning efforttranscribed from text Source excerpt Terminal-Bench 4.0 — Accuracy vs Cost: "GPT-5.6 Sol · Low: 7.9% $1.46"
Terminal-Bench 4.0Low	Low	7.9%	$1.46Cost per attempt	OpenAI	Introducing Claude Sonnet 5.5official · Retrieved 2026-11-12	Vendor-reported value. A point in the figure "Terminal-Bench 4.0 / Accuracy vs. cost" on the Anthropic announcement page. values were taken from each point's aria-label (DOM); the page shows them only in the figure; value for a model from some other vendor (announced by Anthropic); the figure's effort label "Low" was mapped to low in the OpenAI API reasoning effort; measured by: the Terminal-Bench 4.0 table footnote on the duplicate row for the same point (Opus 5.5 announcement page) states "GPT-6 Astra and GPT-5.6 Sol figures are as reported by OpenAI"transcribed from text Source excerpt Terminal-Bench 4.0 — Accuracy vs. cost: "GPT-5.6 Sol · Low: 7.9% $1.46"
Terminal-Bench 4.0Med (default)	Med	20.9%	$2.69Cost per attempt	OpenAI	Introducing Claude Opus 5.5official · Retrieved 2026-11-12	Evidence (same value as comparison group Accuracy vs Cost figure announced by Anthropic (Opus 5.5 and Sonnet 5.5 pages))Vendor-reported value. A point in the figure "Terminal-Bench 4.0 / Accuracy vs Cost" on the Anthropic announcement page. values were taken from each point's aria-label (DOM); the page shows them only in the figure; value for a model from some other vendor. measured by: the figure makes no distinction / the table footnote says GPT series figures are reported values from OpenAI; the figure's effort label "Med" was mapped to medium in the OpenAI API reasoning efforttranscribed from text Source excerpt Terminal-Bench 4.0 — Accuracy vs Cost: "GPT-5.6 Sol · Med: 20.9% $2.69"
Terminal-Bench 4.0Med (default)	Med	20.9%	$2.69Cost per attempt	OpenAI	Introducing Claude Sonnet 5.5official · Retrieved 2026-11-12	Vendor-reported value. A point in the figure "Terminal-Bench 4.0 / Accuracy vs. cost" on the Anthropic announcement page. values were taken from each point's aria-label (DOM); the page shows them only in the figure; value for a model from some other vendor (announced by Anthropic); the figure's effort label "Med" was mapped to medium in the OpenAI API reasoning effort; measured by: the Terminal-Bench 4.0 table footnote on the duplicate row for the same point (Opus 5.5 announcement page) states "GPT-6 Astra and GPT-5.6 Sol figures are as reported by OpenAI"transcribed from text Source excerpt Terminal-Bench 4.0 — Accuracy vs. cost: "GPT-5.6 Sol · Med: 20.9% $2.69"
Terminal-Bench 4.0High	High	26.1%	$4.12Cost per attempt	OpenAI	Introducing Claude Opus 5.5official · Retrieved 2026-11-12	Evidence (same value as comparison group Accuracy vs Cost figure announced by Anthropic (Opus 5.5 and Sonnet 5.5 pages))Vendor-reported value. A point in the figure "Terminal-Bench 4.0 / Accuracy vs Cost" on the Anthropic announcement page. values were taken from each point's aria-label (DOM); the page shows them only in the figure; value for a model from some other vendor. measured by: the figure makes no distinction / the table footnote says GPT series figures are reported values from OpenAI; the figure's effort label "High" was mapped to high in the OpenAI API reasoning efforttranscribed from text Source excerpt Terminal-Bench 4.0 — Accuracy vs Cost: "GPT-5.6 Sol · High: 26.1% $4.12"
Terminal-Bench 4.0High	High	26.1%	$4.12Cost per attempt	OpenAI	Introducing Claude Sonnet 5.5official · Retrieved 2026-11-12	Vendor-reported value. A point in the figure "Terminal-Bench 4.0 / Accuracy vs. cost" on the Anthropic announcement page. values were taken from each point's aria-label (DOM); the page shows them only in the figure; value for a model from some other vendor (announced by Anthropic); the figure's effort label "High" was mapped to high in the OpenAI API reasoning effort; measured by: the Terminal-Bench 4.0 table footnote on the duplicate row for the same point (Opus 5.5 announcement page) states "GPT-6 Astra and GPT-5.6 Sol figures are as reported by OpenAI"transcribed from text Source excerpt Terminal-Bench 4.0 — Accuracy vs. cost: "GPT-5.6 Sol · High: 26.1% $4.12"
Terminal-Bench 4.0Xhigh	Xhigh	28.5%	$5.39Cost per attempt	OpenAI	Introducing Claude Opus 5.5official · Retrieved 2026-11-12	Evidence (same value as comparison group Accuracy vs Cost figure announced by Anthropic (Opus 5.5 and Sonnet 5.5 pages))Vendor-reported value. A point in the figure "Terminal-Bench 4.0 / Accuracy vs Cost" on the Anthropic announcement page. values were taken from each point's aria-label (DOM); the page shows them only in the figure; value for a model from some other vendor. measured by: the figure makes no distinction / the table footnote says GPT series figures are reported values from OpenAI; the figure's effort label "Xhigh" was mapped to xhigh in the OpenAI API reasoning efforttranscribed from text Source excerpt Terminal-Bench 4.0 — Accuracy vs Cost: "GPT-5.6 Sol · Xhigh: 28.5% $5.39"
Terminal-Bench 4.0Xhigh	Xhigh	28.5%	$5.39Cost per attempt	OpenAI	Introducing Claude Sonnet 5.5official · Retrieved 2026-11-12	Vendor-reported value. A point in the figure "Terminal-Bench 4.0 / Accuracy vs. cost" on the Anthropic announcement page. values were taken from each point's aria-label (DOM); the page shows them only in the figure; value for a model from some other vendor (announced by Anthropic); the figure's effort label "Xhigh" was mapped to xhigh in the OpenAI API reasoning effort; measured by: the Terminal-Bench 4.0 table footnote on the duplicate row for the same point (Opus 5.5 announcement page) states "GPT-6 Astra and GPT-5.6 Sol figures are as reported by OpenAI"transcribed from text Source excerpt Terminal-Bench 4.0 — Accuracy vs. cost: "GPT-5.6 Sol · Xhigh: 28.5% $5.39"
Terminal-Bench 4.0Max	Max	37.3%	$7.89Cost per attempt	OpenAI	Introducing Claude Opus 5.5official · Retrieved 2026-11-12	Evidence (same value as comparison group Accuracy vs Cost figure announced by Anthropic (Opus 5.5 and Sonnet 5.5 pages))Vendor-reported value. A point in the figure "Terminal-Bench 4.0 / Accuracy vs Cost" on the Anthropic announcement page. values were taken from each point's aria-label (DOM); the page shows them only in the figure; value for a model from some other vendor. measured by: the figure makes no distinction / the table footnote says GPT series figures are reported values from OpenAI; the figure's effort label "Max" was mapped to max in the OpenAI API reasoning efforttranscribed from text Source excerpt Terminal-Bench 4.0 — Accuracy vs Cost: "GPT-5.6 Sol · Max: 37.3% $7.89"
Terminal-Bench 4.0Max	Max	37.3%	$7.89Cost per attempt	OpenAI	Introducing Claude Sonnet 5.5official · Retrieved 2026-11-12	Vendor-reported value. A point in the figure "Terminal-Bench 4.0 / Accuracy vs. cost" on the Anthropic announcement page. values were taken from each point's aria-label (DOM); the page shows them only in the figure; value for a model from some other vendor (announced by Anthropic); the figure's effort label "Max" was mapped to max in the OpenAI API reasoning effort; measured by: the Terminal-Bench 4.0 table footnote on the duplicate row for the same point (Opus 5.5 announcement page) states "GPT-6 Astra and GPT-5.6 Sol figures are as reported by OpenAI"transcribed from text Source excerpt Terminal-Bench 4.0 — Accuracy vs. cost: "GPT-5.6 Sol · Max: 37.3% $7.89"
Terminal-Bench-Science 0.1not stated	not stated	22.4%	—	not stated	Introducing Claude Opus 5.5official · Retrieved 2026-11-12	Vendor-reported value. table note: Unless otherwise noted, all Claude Opus 5.5 results use adaptive thinking at max effort. production safeguards enabled. Values for other vendor models (as listed by Anthropic; null where effort was not stated). SE ±3.5–5pt. The GPT-6 Astra figure is a reported value from OpenAI.transcribed from table Source excerpt 22.4%Location Table, Terminal-Bench-Science 0.1 row: 22.4
FrontierCode v1.1(main set)Low	Low	35.4%	$1.75Cost per task	not stated	Introducing Claude Opus 5.5official · Retrieved 2026-11-12	Vendor-reported value. A point in the figure "FrontierCode v1.1, main set / Accuracy vs Cost" on the Anthropic announcement page. values were taken from each point's aria-label (DOM); the page shows them only in the figure; value for a model from some other vendor (announced by Anthropic); the figure's effort label "Low" was mapped to low in the OpenAI API reasoning efforttranscribed from text Source excerpt FrontierCode v1.1, main set — Accuracy vs Cost: "GPT-5.6 Sol · Low: 35.4% $1.75"
FrontierCode v1.1(main set)Med (default)	Med	39.9%	$2.50Cost per task	not stated	Introducing Claude Opus 5.5official · Retrieved 2026-11-12	Vendor-reported value. A point in the figure "FrontierCode v1.1, main set / Accuracy vs Cost" on the Anthropic announcement page. values were taken from each point's aria-label (DOM); the page shows them only in the figure; value for a model from some other vendor (announced by Anthropic); the figure's effort label "Med" was mapped to medium in the OpenAI API reasoning efforttranscribed from text Source excerpt FrontierCode v1.1, main set — Accuracy vs Cost: "GPT-5.6 Sol · Med: 39.9% $2.50"
FrontierCode v1.1(main set)High	High	45.1%	$3.25Cost per task	not stated	Introducing Claude Opus 5.5official · Retrieved 2026-11-12	Vendor-reported value. A point in the figure "FrontierCode v1.1, main set / Accuracy vs Cost" on the Anthropic announcement page. values were taken from each point's aria-label (DOM); the page shows them only in the figure; value for a model from some other vendor (announced by Anthropic); the figure's effort label "High" was mapped to high in the OpenAI API reasoning efforttranscribed from text Source excerpt FrontierCode v1.1, main set — Accuracy vs Cost: "GPT-5.6 Sol · High: 45.1% $3.25"
FrontierCode v1.1(main set)Xhigh	Xhigh	46.8%	$3.88Cost per task	not stated	Introducing Claude Opus 5.5official · Retrieved 2026-11-12	Vendor-reported value. A point in the figure "FrontierCode v1.1, main set / Accuracy vs Cost" on the Anthropic announcement page. values were taken from each point's aria-label (DOM); the page shows them only in the figure; value for a model from some other vendor (announced by Anthropic); the figure's effort label "Xhigh" was mapped to xhigh in the OpenAI API reasoning efforttranscribed from text Source excerpt FrontierCode v1.1, main set — Accuracy vs Cost: "GPT-5.6 Sol · Xhigh: 46.8% $3.88"
FrontierCode v1.1(main set)Max	Max	47.5%	$4.85Cost per task	not stated	Introducing Claude Opus 5.5official · Retrieved 2026-11-12	Vendor-reported value. A point in the figure "FrontierCode v1.1, main set / Accuracy vs Cost" on the Anthropic announcement page. values were taken from each point's aria-label (DOM); the page shows them only in the figure; value for a model from some other vendor (announced by Anthropic); the figure's effort label "Max" was mapped to max in the OpenAI API reasoning efforttranscribed from text Source excerpt FrontierCode v1.1, main set — Accuracy vs Cost: "GPT-5.6 Sol · Max: 47.5% $4.85"
CursorBench 4.0Low	Low	24.6%	$0.87Cost per task	not stated	Introducing Claude Opus 5.5official · Retrieved 2026-11-12	Vendor-reported value. A point in the figure "CursorBench 4.0 / Accuracy vs Cost" on the Anthropic announcement page. values were taken from each point's aria-label (DOM); the page shows them only in the figure; value for a model from some other vendor (announced by Anthropic); the figure's effort label "Low" was mapped to low in the OpenAI API reasoning efforttranscribed from text Source excerpt CursorBench 4.0 — Accuracy vs Cost: "GPT-5.6 Sol · Low: 24.6% $0.87"
CursorBench 4.0Low	Low	24.6%	$0.87Cost per task	not stated	Introducing Claude Sonnet 5.5official · Retrieved 2026-11-12	Vendor-reported value. A point in the figure "CursorBench 4.0 / Accuracy vs. cost" on the Anthropic announcement page. values were taken from each point's aria-label (DOM); the page shows them only in the figure; value for a model from some other vendor (announced by Anthropic); the figure's effort label "Low" was mapped to low in the OpenAI API reasoning efforttranscribed from text Source excerpt CursorBench 4.0 — Accuracy vs. cost: "GPT-5.6 Sol · Low: 24.6% $0.87"
CursorBench 4.0Med (default)	Med	31.1%	$1.77Cost per task	not stated	Introducing Claude Opus 5.5official · Retrieved 2026-11-12	Vendor-reported value. A point in the figure "CursorBench 4.0 / Accuracy vs Cost" on the Anthropic announcement page. values were taken from each point's aria-label (DOM); the page shows them only in the figure; value for a model from some other vendor (announced by Anthropic); the figure's effort label "Med" was mapped to medium in the OpenAI API reasoning efforttranscribed from text Source excerpt CursorBench 4.0 — Accuracy vs Cost: "GPT-5.6 Sol · Med: 31.1% $1.77"
CursorBench 4.0Med (default)	Med	31.1%	$1.77Cost per task	not stated	Introducing Claude Sonnet 5.5official · Retrieved 2026-11-12	Vendor-reported value. A point in the figure "CursorBench 4.0 / Accuracy vs. cost" on the Anthropic announcement page. values were taken from each point's aria-label (DOM); the page shows them only in the figure; value for a model from some other vendor (announced by Anthropic); the figure's effort label "Med" was mapped to medium in the OpenAI API reasoning efforttranscribed from text Source excerpt CursorBench 4.0 — Accuracy vs. cost: "GPT-5.6 Sol · Med: 31.1% $1.77"
CursorBench 4.0High	High	35.7%	$2.85Cost per task	not stated	Introducing Claude Opus 5.5official · Retrieved 2026-11-12	Vendor-reported value. A point in the figure "CursorBench 4.0 / Accuracy vs Cost" on the Anthropic announcement page. values were taken from each point's aria-label (DOM); the page shows them only in the figure; value for a model from some other vendor (announced by Anthropic); the figure's effort label "High" was mapped to high in the OpenAI API reasoning efforttranscribed from text Source excerpt CursorBench 4.0 — Accuracy vs Cost: "GPT-5.6 Sol · High: 35.7% $2.85"
CursorBench 4.0High	High	35.7%	$2.85Cost per task	not stated	Introducing Claude Sonnet 5.5official · Retrieved 2026-11-12	Vendor-reported value. A point in the figure "CursorBench 4.0 / Accuracy vs. cost" on the Anthropic announcement page. values were taken from each point's aria-label (DOM); the page shows them only in the figure; value for a model from some other vendor (announced by Anthropic); the figure's effort label "High" was mapped to high in the OpenAI API reasoning efforttranscribed from text Source excerpt CursorBench 4.0 — Accuracy vs. cost: "GPT-5.6 Sol · High: 35.7% $2.85"
CursorBench 4.0Xhigh	Xhigh	37.7%	$4.40Cost per task	not stated	Introducing Claude Opus 5.5official · Retrieved 2026-11-12	Vendor-reported value. A point in the figure "CursorBench 4.0 / Accuracy vs Cost" on the Anthropic announcement page. values were taken from each point's aria-label (DOM); the page shows them only in the figure; value for a model from some other vendor (announced by Anthropic); the figure's effort label "Xhigh" was mapped to xhigh in the OpenAI API reasoning efforttranscribed from text Source excerpt CursorBench 4.0 — Accuracy vs Cost: "GPT-5.6 Sol · Xhigh: 37.7% $4.40"
CursorBench 4.0Xhigh	Xhigh	37.7%	$4.40Cost per task	not stated	Introducing Claude Sonnet 5.5official · Retrieved 2026-11-12	Vendor-reported value. A point in the figure "CursorBench 4.0 / Accuracy vs. cost" on the Anthropic announcement page. values were taken from each point's aria-label (DOM); the page shows them only in the figure; value for a model from some other vendor (announced by Anthropic); the figure's effort label "Xhigh" was mapped to xhigh in the OpenAI API reasoning efforttranscribed from text Source excerpt CursorBench 4.0 — Accuracy vs. cost: "GPT-5.6 Sol · Xhigh: 37.7% $4.40"
CursorBench 4.0Max	Max	41.7%	$8.23Cost per task	not stated	Introducing Claude Opus 5.5official · Retrieved 2026-11-12	Vendor-reported value. A point in the figure "CursorBench 4.0 / Accuracy vs Cost" on the Anthropic announcement page. values were taken from each point's aria-label (DOM); the page shows them only in the figure; value for a model from some other vendor (announced by Anthropic); the figure's effort label "Max" was mapped to max in the OpenAI API reasoning efforttranscribed from text Source excerpt CursorBench 4.0 — Accuracy vs Cost: "GPT-5.6 Sol · Max: 41.7% $8.23"
CursorBench 4.0Max	Max	41.7%	$8.23Cost per task	not stated	Introducing Claude Sonnet 5.5official · Retrieved 2026-11-12	Vendor-reported value. A point in the figure "CursorBench 4.0 / Accuracy vs. cost" on the Anthropic announcement page. values were taken from each point's aria-label (DOM); the page shows them only in the figure; value for a model from some other vendor (announced by Anthropic); the figure's effort label "Max" was mapped to max in the OpenAI API reasoning efforttranscribed from text Source excerpt CursorBench 4.0 — Accuracy vs. cost: "GPT-5.6 Sol · Max: 41.7% $8.23"
GDPval-AA v2.1Low	Low	1,289	$0.27Estimated cost per task	not stated	Introducing Claude Opus 5.5official · Retrieved 2026-11-12	Vendor-reported value. A point in the figure "GDPval-AA v2.1 / Elo vs Cost" on the Anthropic announcement page. values were taken from each point's aria-label (DOM); the page shows them only in the figure; value for a model from some other vendor (announced by Anthropic); the figure's effort label "Low" was mapped to low in the OpenAI API reasoning efforttranscribed from text Source excerpt GDPval-AA v2.1 — Elo vs Cost: "GPT-5.6 Sol · Low: 1289 $0.27"
GDPval-AA v2.1Med (default)	Med	1,403	$0.60Estimated cost per task	not stated	Introducing Claude Opus 5.5official · Retrieved 2026-11-12	Vendor-reported value. A point in the figure "GDPval-AA v2.1 / Elo vs Cost" on the Anthropic announcement page. values were taken from each point's aria-label (DOM); the page shows them only in the figure; value for a model from some other vendor (announced by Anthropic); the figure's effort label "Med" was mapped to medium in the OpenAI API reasoning efforttranscribed from text Source excerpt GDPval-AA v2.1 — Elo vs Cost: "GPT-5.6 Sol · Med: 1403 $0.60"
GDPval-AA v2.1High	High	1,480	$1.11Estimated cost per task	not stated	Introducing Claude Opus 5.5official · Retrieved 2026-11-12	Vendor-reported value. A point in the figure "GDPval-AA v2.1 / Elo vs Cost" on the Anthropic announcement page. values were taken from each point's aria-label (DOM); the page shows them only in the figure; value for a model from some other vendor (announced by Anthropic); the figure's effort label "High" was mapped to high in the OpenAI API reasoning efforttranscribed from text Source excerpt GDPval-AA v2.1 — Elo vs Cost: "GPT-5.6 Sol · High: 1480 $1.11"
GDPval-AA v2.1Xhigh	Xhigh	1,548	$1.70Estimated cost per task	not stated	Introducing Claude Opus 5.5official · Retrieved 2026-11-12	Vendor-reported value. A point in the figure "GDPval-AA v2.1 / Elo vs Cost" on the Anthropic announcement page. values were taken from each point's aria-label (DOM); the page shows them only in the figure; value for a model from some other vendor (announced by Anthropic); the figure's effort label "Xhigh" was mapped to xhigh in the OpenAI API reasoning efforttranscribed from text Source excerpt GDPval-AA v2.1 — Elo vs Cost: "GPT-5.6 Sol · Xhigh: 1548 $1.70"
GDPval-AA v2.1Max	Max	1,588	$2.81Estimated cost per task	not stated	Introducing Claude Opus 5.5official · Retrieved 2026-11-12	Vendor-reported value. A point in the figure "GDPval-AA v2.1 / Elo vs Cost" on the Anthropic announcement page. values were taken from each point's aria-label (DOM); the page shows them only in the figure; value for a model from some other vendor (announced by Anthropic); the figure's effort label "Max" was mapped to max in the OpenAI API reasoning efforttranscribed from text Source excerpt GDPval-AA v2.1 — Elo vs Cost: "GPT-5.6 Sol · Max: 1588 $2.81"
AutomationBenchLow	Low	11.7%	$0.42Cost per task	Zapier	Introducing Claude Opus 5.5official · Retrieved 2026-11-12	Vendor-reported value. A point in the figure "AutomationBench / Accuracy vs Cost" on the Anthropic announcement page. values were taken from each point's aria-label (DOM); the page shows them only in the figure; measured by: Zapier (Opus 5.5 is from a Zapier early access evaluation; the others are from the Zapier public leaderboard); the figure's effort label "Low" was mapped to low in the OpenAI API reasoning efforttranscribed from text Source excerpt AutomationBench — Accuracy vs Cost: "GPT-5.6 Sol · Low: 11.7% $0.42"
AutomationBenchMed (default)	Med	19.6%	$0.57Cost per task	Zapier	Introducing Claude Opus 5.5official · Retrieved 2026-11-12	Vendor-reported value. A point in the figure "AutomationBench / Accuracy vs Cost" on the Anthropic announcement page. values were taken from each point's aria-label (DOM); the page shows them only in the figure; measured by: Zapier (Opus 5.5 is from a Zapier early access evaluation; the others are from the Zapier public leaderboard); the figure's effort label "Med" was mapped to medium in the OpenAI API reasoning efforttranscribed from text Source excerpt AutomationBench — Accuracy vs Cost: "GPT-5.6 Sol · Med: 19.6% $0.57"
AutomationBenchHigh	High	24.8%	$0.64Cost per task	Zapier	Introducing Claude Opus 5.5official · Retrieved 2026-11-12	Vendor-reported value. A point in the figure "AutomationBench / Accuracy vs Cost" on the Anthropic announcement page. values were taken from each point's aria-label (DOM); the page shows them only in the figure; measured by: Zapier (Opus 5.5 is from a Zapier early access evaluation; the others are from the Zapier public leaderboard); the figure's effort label "High" was mapped to high in the OpenAI API reasoning efforttranscribed from text Source excerpt AutomationBench — Accuracy vs Cost: "GPT-5.6 Sol · High: 24.8% $0.64"
AutomationBenchXhigh	Xhigh	26.3%	$0.74Cost per task	Zapier	Introducing Claude Opus 5.5official · Retrieved 2026-11-12	Vendor-reported value. A point in the figure "AutomationBench / Accuracy vs Cost" on the Anthropic announcement page. values were taken from each point's aria-label (DOM); the page shows them only in the figure; measured by: Zapier (Opus 5.5 is from a Zapier early access evaluation; the others are from the Zapier public leaderboard); the figure's effort label "Xhigh" was mapped to xhigh in the OpenAI API reasoning efforttranscribed from text Source excerpt AutomationBench — Accuracy vs Cost: "GPT-5.6 Sol · Xhigh: 26.3% $0.74"
AutomationBenchMax	Max	28.8%	$0.91Cost per task	Zapier	Introducing Claude Opus 5.5official · Retrieved 2026-11-12	Vendor-reported value. A point in the figure "AutomationBench / Accuracy vs Cost" on the Anthropic announcement page. values were taken from each point's aria-label (DOM); the page shows them only in the figure; measured by: Zapier (Opus 5.5 is from a Zapier early access evaluation; the others are from the Zapier public leaderboard); the figure's effort label "Max" was mapped to max in the OpenAI API reasoning efforttranscribed from text Source excerpt AutomationBench — Accuracy vs Cost: "GPT-5.6 Sol · Max: 28.8% $0.91"
HealthBench Professional (length-adjusted score)not stated	not stated	60.5	—	OpenAI	GPT-6 Astra System Card (Table 6. HealthBench Evaluations)official · Retrieved 2026-11-12	Vendor-reported value. Self-reported by OpenAI. effort not stated. length-adjusted scoretranscribed from table Source excerpt 60.5Location Table 6 "HealthBench Evaluations - reported as length-adjusted score", HealthBench Professional row
HealthBench Hard (length-adjusted score)not stated	not stated	33.1	—	OpenAI	GPT-6 Astra System Card (Table 6. HealthBench Evaluations)official · Retrieved 2026-11-12	Vendor-reported value. Self-reported by OpenAI. effort not stated. length-adjusted scoretranscribed from table Source excerpt 33.1Location Table 6 "HealthBench Evaluations - reported as length-adjusted score", HealthBench Hard row
GPQA DiamondNone	None	82.8%	—	Epoch AI	Epoch AI, Capabilities & benchmarking (benchmark_data.zip, gpqa_diamond.csv) CC BY 4.0third-party · Retrieved 2026-11-12	Run by Epoch AI (third-party measurement). Source label "GPT-5.6 Sol (none)". effort is indicated by the suffix _none in Epoch's Model version. Data retrieved 2026-11-12.transcribed from table Source excerpt gpqa_diamond.csv: Model version=gpt-5.6-sol_none, Best score (across scorers)=0.8282828282828283, stderr=0.026869716187429872, Started at=2026-08-07T01:00:09.000ZLocation gpqa_diamond.csv, data row 83 (file line 136), id=DDfYPyKQW7vgDtGgkwcKSi, Model version=gpt-5.6-sol_none
GPQA DiamondLow	Low	89.9%	—	Epoch AI	Epoch AI, Capabilities & benchmarking (benchmark_data.zip, gpqa_diamond.csv) CC BY 4.0third-party · Retrieved 2026-11-12	Run by Epoch AI (third-party measurement). Source label "GPT-5.6 Sol (low)". effort is indicated by the suffix _low in Epoch's Model version. Data retrieved 2026-11-12.transcribed from table Source excerpt gpqa_diamond.csv: Model version=gpt-5.6-sol_low, Best score (across scorers)=0.898989898989899, stderr=0.021469735576055356, Started at=2026-08-07T00:46:32.000ZLocation gpqa_diamond.csv, data row 90 (file line 147), id=6cmgkNAQZekBr6e6rVdLe9, Model version=gpt-5.6-sol_low
GPQA DiamondMax	Max	93.5%	—	Epoch AI	Epoch AI, Capabilities & benchmarking (benchmark_data.zip, gpqa_diamond.csv) CC BY 4.0third-party · Retrieved 2026-11-12	Run by Epoch AI (third-party measurement). Source label "GPT-5.6 Sol (max)". effort is indicated by the suffix _max in Epoch's Model version. Data retrieved 2026-11-12.transcribed from table Source excerpt gpqa_diamond.csv: Model version=gpt-5.6-sol_max, Best score (across scorers)=0.9349747474747475, stderr=0.01570109070867593, Started at=2026-07-09T02:07:13.000ZLocation gpqa_diamond.csv, data row 170 (file line 240), id=nuKCgZReCBsXKCkNj9bK37, Model version=gpt-5.6-sol_max
FrontierMath Tiers 1-3(v2)Max	Max	89.1%	—	Epoch AI	Epoch AI, Capabilities & benchmarking (benchmark_data.zip, frontiermath_tiers_1_3_v2.csv) CC BY 4.0third-party · Retrieved 2026-11-12	Run by Epoch AI (third-party measurement). Source label "GPT-5.6 Sol (max)". effort is indicated by the suffix _max in Epoch's Model version. Data retrieved 2026-11-12.transcribed from table Source excerpt frontiermath_tiers_1_3_v2.csv: Model version=gpt-5.6-sol_max, Best score (across scorers)=0.8912280701754386, stderr=0.018475392571140503, Started at=2026-07-09T02:45:09.000ZLocation frontiermath_tiers_1_3_v2.csv, data row 92 (file line 122), id=VBpT2WWYKBZtUxG9Z3svGh, Model version=gpt-5.6-sol_max
FrontierMath Tier 4(v2)Max	Max	82.9%	—	Epoch AI	Epoch AI, Capabilities & benchmarking (benchmark_data.zip, frontiermath_tier_4_v2.csv) CC BY 4.0third-party · Retrieved 2026-11-12	Run by Epoch AI (third-party measurement). Source label "GPT-5.6 Sol (max)". effort is indicated by the suffix _max in Epoch's Model version. Data retrieved 2026-11-12.transcribed from table Source excerpt frontiermath_tier_4_v2.csv: Model version=gpt-5.6-sol_max, Best score (across scorers)=0.8292682926829268, stderr=0.05949419959829497, Started at=2026-07-09T02:45:09.000ZLocation frontiermath_tier_4_v2.csv, data row 47 (file line 84), id=QJ5rJMVypB4PMxPcBGftc8, Model version=gpt-5.6-sol_max
OTIS Mock AIME 2024-2025None	None	68.9%	—	Epoch AI	Epoch AI, Capabilities & benchmarking (benchmark_data.zip, otis_mock_aime_2024_2025.csv) CC BY 4.0third-party · Retrieved 2026-11-12	Run by Epoch AI (third-party measurement). Source label "GPT-5.6 Sol (none)". effort is indicated by the suffix _none in Epoch's Model version. Data retrieved 2026-11-12.transcribed from table Source excerpt otis_mock_aime_2024_2025.csv: Model version=gpt-5.6-sol_none, Best score (across scorers)=0.6888888888888889, stderr=0.0697920592732311, Started at=2026-08-07T00:54:12.000ZLocation otis_mock_aime_2024_2025.csv, data row 91 (file line 151), id=g7czyLHudHegzP9yRD939X, Model version=gpt-5.6-sol_none
OTIS Mock AIME 2024-2025Low	Low	95.6%	—	Epoch AI	Epoch AI, Capabilities & benchmarking (benchmark_data.zip, otis_mock_aime_2024_2025.csv) CC BY 4.0third-party · Retrieved 2026-11-12	Run by Epoch AI (third-party measurement). Source label "GPT-5.6 Sol (low)". effort is indicated by the suffix _low in Epoch's Model version. Data retrieved 2026-11-12.transcribed from table Source excerpt otis_mock_aime_2024_2025.csv: Model version=gpt-5.6-sol_low, Best score (across scorers)=0.9555555555555556, stderr=0.031067790907534736, Started at=2026-08-07T00:42:21.000ZLocation otis_mock_aime_2024_2025.csv, data row 98 (file line 162), id=Mifz96Jm442bn43Qnxj5rT, Model version=gpt-5.6-sol_low
OTIS Mock AIME 2024-2025Max	Max	100.0%	—	Epoch AI	Epoch AI, Capabilities & benchmarking (benchmark_data.zip, otis_mock_aime_2024_2025.csv) CC BY 4.0third-party · Retrieved 2026-11-12	Run by Epoch AI (third-party measurement). Source label "GPT-5.6 Sol (max)". effort is indicated by the suffix _max in Epoch's Model version. Data retrieved 2026-11-12.transcribed from table Source excerpt otis_mock_aime_2024_2025.csv: Model version=gpt-5.6-sol_max, Best score (across scorers)=1.0, stderr=0.0, Started at=2026-07-09T02:07:13.000ZLocation otis_mock_aime_2024_2025.csv, data row 178 (file line 253), id=7RjsANKwwsePSstcZeSMhX, Model version=gpt-5.6-sol_max
SimpleQA VerifiedMax	Max	69.7%	—	Epoch AI	Epoch AI, Capabilities & benchmarking (benchmark_data.zip, simpleqa_verified.csv) CC BY 4.0third-party · Retrieved 2026-11-12	Run by Epoch AI (third-party measurement). Source label "GPT-5.6 Sol (max)". effort is indicated by the suffix _max in Epoch's Model version. Data retrieved 2026-11-12.transcribed from table Source excerpt simpleqa_verified.csv: Model version=gpt-5.6-sol_max, Best score (across scorers)=0.697, stderr=0.014539683710535193, Started at=2026-08-10T23:44:30.000ZLocation simpleqa_verified.csv, data row 92 (file line 142), id=UnRUSuZzm2X27TUbpeMPva, Model version=gpt-5.6-sol_max
Chess Puzzles(Epoch AI)None	None	7.0%	—	Epoch AI	Epoch AI, Capabilities & benchmarking (benchmark_data.zip, chess_puzzles.csv) CC BY 4.0third-party · Retrieved 2026-11-12	Run by Epoch AI (third-party measurement). Source label "GPT-5.6 Sol (none)". effort is indicated by the suffix _none in Epoch's Model version. Data retrieved 2026-11-12.transcribed from table Source excerpt chess_puzzles.csv: Model version=gpt-5.6-sol_none, Best score (across scorers)=0.07, stderr=0.02564323999762428, Started at=2026-10-13T13:49:41.000ZLocation chess_puzzles.csv, data row 90 (file line 141), id=WvV6QREDDXeDEGdrKztDsy, Model version=gpt-5.6-sol_none
Chess Puzzles(Epoch AI)Low	Low	31.0%	—	Epoch AI	Epoch AI, Capabilities & benchmarking (benchmark_data.zip, chess_puzzles.csv) CC BY 4.0third-party · Retrieved 2026-11-12	Run by Epoch AI (third-party measurement). Source label "GPT-5.6 Sol (low)". effort is indicated by the suffix _low in Epoch's Model version. Data retrieved 2026-11-12.transcribed from table Source excerpt chess_puzzles.csv: Model version=gpt-5.6-sol_low, Best score (across scorers)=0.31, stderr=0.04648231987117317, Started at=2026-10-13T13:49:48.000ZLocation chess_puzzles.csv, data row 89 (file line 140), id=V6aVGtn9gAmRep2oBGsqVN, Model version=gpt-5.6-sol_low
Chess Puzzles(Epoch AI)Max	Max	59.0%	—	Epoch AI	Epoch AI, Capabilities & benchmarking (benchmark_data.zip, chess_puzzles.csv) CC BY 4.0third-party · Retrieved 2026-11-12	Run by Epoch AI (third-party measurement). Source label "GPT-5.6 Sol (max)". effort is indicated by the suffix _max in Epoch's Model version. Data retrieved 2026-11-12.transcribed from table Source excerpt chess_puzzles.csv: Model version=gpt-5.6-sol_max, Best score (across scorers)=0.59, stderr=0.049431107042371025, Started at=2026-10-13T13:50:12.000ZLocation chess_puzzles.csv, data row 88 (file line 139), id=oWqwMmwEd7LU3eU5Ftak64, Model version=gpt-5.6-sol_max
MirrorCodeHigh	High	20.0%	—	Epoch AI	Epoch AI, Capabilities & benchmarking (benchmark_data.zip, mirrorcode.csv) CC BY 4.0third-party · Retrieved 2026-11-12	Run by Epoch AI (third-party measurement). Source label "GPT-5.6 Sol (high)". effort is indicated by the suffix _high in Epoch's Model version. Data retrieved 2026-11-12.transcribed from table Source excerpt mirrorcode.csv: Model version=gpt-5.6-sol_high, Best score (across scorers)=0.2, stderr=0.09229582069908972, Started at=2026-08-10T16:12:44.772ZLocation mirrorcode.csv, data row 44 (file line 67), id=mirrorcode-ml-2l-gpt-5-6-sol, Model version=gpt-5.6-sol_high
Epoch Capabilities Index(ECI)not stated	not stated	162.99	—	Epoch AI	Epoch AI, epoch_capabilities_index/eci_scores.csv (CC BY 4.0)third-party · Retrieved 2026-11-12	A per-model metric that combines results across multiple benchmarks and multiple efforts (not per effort). Calculated by Epoch AI.transcribed from table Source excerpt eci_scores.csv: Model=GPT-5.6 Sol, eci=162.99, eci_ci_low=160.68, eci_ci_high=165.49Location epoch_capabilities_index/eci_scores.csv, data row 13 (file line 14), Model=GPT-5.6 Sol
τ²-bench banking_knowledge (v1.0.1)Xhigh	Xhigh	46.9%	—	Sierra	τ²-bench leaderboard submission gpt-5-6-sol_sierra_2026-08-04 (sierra-research/tau2-bench, MIT)third-party · Retrieved 2026-11-13	τ²-bench leaderboard submission gpt-5-6-sol_sierra_2026-08-04 (submission_type standard, tau2_bench_version 1.0.1). Submitter "Sierra", model_organization "OpenAI": Run by Sierra, which operates τ²-bench (third-party measurement). Verified on the leaderboard (trajectories available, prompts not modified, no questions omitted). The value is pass^1 (%). reasoning_effort "xhigh". User simulator gpt-5.2 (reasoning_effort low). Search configuration retrieval_config "alltools". Evaluation date 2026-07-22. Notes summary: 4 trial runs; seed 300.transcribed from table Source excerpt banking_knowledge.pass_1=46.90721649484536Location In submissions/gpt-5-6-sol_sierra_2026-08-04/submission.json: results.banking_knowledge.pass_1 (line 17 of file). model_name "GPT-5.6-sol". methodology.notes (line 36)

GPT-5.6 Sol

Model information

Price (USD per 1 million tokens)

Benchmark results

Related rankings

References