# GPT-6 AstraHigh

> Source: Benchize (https://benchize.com) — https://benchize.com/models/gpt-6-astra-high. When citing or reusing this content, credit "Benchize" and link to https://benchize.com/models/gpt-6-astra-high.

OpenAI

Canonical URL: https://benchize.com/models/gpt-6-astra-high
Last updated: Thu, 22 Oct 2026 05:40:11 GMT

Reproduction and quotation must name Benchize, link to the canonical HTML URL, and retain individual source citations.

OpenAI

## Benchmarks

| Benchmark | Score | ± | Rank | Cost | Latency | Source |  |
| --- | --- | --- | --- | --- | --- | --- | --- |
| [MirrorCode](https://benchize.com/benchmarks/epoch-ai%3Amirrorcode.md) | 46.7% | SE ±11.86 pp | 5 / 9 | — | — | [epoch ai · MirrorCode](https://epoch.ai/data/benchmark_data.zip#file=mirrorcode.csv) |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `score` \| 0.467 \| score_fraction \| Best score (across scorers) \| Higher is better \| \| `mean_score` \| 0.467 \| score_fraction \| mean_score \| Higher is better \| ### Configuration benchmarkVersion MirrorCode reasoningBudget Not reported Reasoning effort high sourceModelVersion gpt-6-astra_high ### Reported statistics Uncertainty Standard error: ±11.86 pp standardError 0.1186 ### Source [Original source](https://epoch.ai/data/benchmark_data.zip#file=mirrorcode.csv) Source epoch ai Original model label gpt-6-astra Source version MirrorCode Evaluated 2026-08-30T16:51:09.000Z Observed 2026-10-22T03:00:05.281Z Evidence SHA-256 `34012d7fe817c85c7baeba0e236d8c163dda4de486cc9d1f5a29c5b6d3196cd9` Source locator `/rows/3` |  |  |  |  |  |  |  |
| [FrontierMath Tier 4 v2 (private)](https://benchize.com/benchmarks/epoch-ai%3Afrontiermath_tier_4_v2.md) | 97.6% | SE ±2.4 pp | 2 / 39 | — | — | [epoch ai · FrontierMath-Tier-4-v2-Private](https://epoch.ai/data/benchmark_data.zip#file=frontiermath_tier_4_v2.csv) |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `score` \| 0.976 \| score_fraction \| Best score (across scorers) \| Higher is better \| \| `mean_score` \| 0.976 \| score_fraction \| mean_score \| Higher is better \| ### Configuration benchmarkVersion FrontierMath-Tier-4-v2-Private reasoningBudget Not reported Reasoning effort high sourceModelVersion gpt-6-astra_high ### Reported statistics Uncertainty Standard error: ±2.4 pp standardError 0.024 ### Source [Original source](https://epoch.ai/data/benchmark_data.zip#file=frontiermath_tier_4_v2.csv) Source epoch ai Original model label gpt-6-astra Source version FrontierMath-Tier-4-v2-Private Evaluated 2026-08-30T02:22:12.000Z Observed 2026-10-22T03:00:05.281Z Evidence SHA-256 `2bbef8e82ac84071e3ce42bf744f4fc39dff04a549ccb1a288e3800918bc2e55` Source locator `/rows/21` |  |  |  |  |  |  |  |
| [Short Story Writing — Pairwise Comparison](https://benchize.com/benchmarks/lechmazur-short-story%3Apairwise.md) | 3.44 | 95% CI 3.37–3.51 | 4 / 50 | — | — | [lechmazur short story · 8ba548d1cffefa5c472d3832e05f64e119a88fa4](https://api.github.com/repos/lechmazur/writing/contents/data/writing_gemini_argon_20261013/ratings.csv?ref=8ba548d1cffefa5c472d3832e05f64e119a88fa4) |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `thurstone_rating` \| 3.4432356224193557 \| index \| pooled_observed_panel \| Higher is better \| \| `win_probability` \| 0.913183161554768 \| score_fraction \| comparison_pool \| Higher is better \| ### Configuration benchmarkScope data/writing_gemini_argon_20261013 Reasoning effort high thinkingEnabled true ### Reported statistics Uncertainty 95% confidence interval: 3.37–3.51 bootstrapSamples 300 confidenceInterval95 ["3.3742839351451726","3.507684861495963"] sourceRank 4 ### Source [Original source](https://api.github.com/repos/lechmazur/writing/contents/data/writing_gemini_argon_20261013/ratings.csv?ref=8ba548d1cffefa5c472d3832e05f64e119a88fa4) Source lechmazur short story Original model label GPT-6 Astra (high) Source version 8ba548d1cffefa5c472d3832e05f64e119a88fa4 Observed 2026-10-21T16:34:21.130Z Evidence SHA-256 `c36baebed9f302ff7ad8fe97907d72e1757a015a39b6a3b23437597cf8e3bff6` Source locator `/rows/3` |  |  |  |  |  |  |  |
| [Terminal-Bench 4](https://benchize.com/benchmarks/harbor%3Aterminal-bench-4.md) | 57.88% | 95% CI 54.91–60.85% | 3 / 17 | — | — | [harbor](https://www.tbench.ai/) |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 57.88 \| percent \| reported \| Higher is better \| \| `pass_at_2` \| 0.6545 \| score_fraction \| reported \| Higher is better \| \| `pass_at_3` \| 0.6864 \| score_fraction \| reported \| Higher is better \| \| `pass_at_4` \| 0.703 \| score_fraction \| reported \| Higher is better \| \| `pass_at_5` \| 0.7121 \| score_fraction \| reported \| Higher is better \| ### Configuration agent Codex Reasoning effort high sourceSubmission 3475050c-bf5e-4261-a3f6-0af5350af13f ### Reported statistics Uncertainty 95% confidence interval: 54.91–60.85% confidenceHalfWidth95 2.97 sampleSize 330 ### Source [Original source](https://www.tbench.ai/) Source harbor Original model label GPT-6 Astra Evaluated 2026-09-03 Observed 2026-10-22T04:00:05.388Z Evidence SHA-256 `f02bf923eb5a5a2fae5b1e9f0a0c9ff5fe0e3779f0324a9a4f0e9ed6e322db82` Source locator `/rows/6` |  |  |  |  |  |  |  |
| [FrontierCode 1.1 extended](https://benchize.com/benchmarks/frontiercode%3Av1_1%3Aextended.md) | 63.07% | — | 12 / 125 | $2.62 | — | [frontiercode](https://cognition.com/data/frontiercode-leaderboard/data.json) |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `score` \| 0.6307 \| score_fraction \| blocking_criteria_adjusted \| Higher is better \| \| `correct` \| 0.6907 \| score_fraction \| raw_rubric \| Higher is better \| \| `cost` \| 2.624 \| USD \| per_rollout \| Lower is better \| ### Configuration harness codex Reasoning effort high sourceVersion v1_1 subset extended ### Reported statistics flaggedRate 0 sampleSize 150 ### Source [Original source](https://cognition.com/data/frontiercode-leaderboard/data.json) Source frontiercode Original model label GPT-6 Astra Observed 2026-10-22T05:40:03.220Z Evidence SHA-256 `ff0040778b7bdf69c705ad7474f771fb5e53cfaf5cf8573b3a4614153543d75f` Source locator `/v1_1/data/GPT-6%20Astra/high/extended` |  |  |  |  |  |  |  |
| [DeepSWE](https://benchize.com/benchmarks/deepswe%3Av1.1.md) | 73.23% | Range 69.81–76.65% | 4 / 70 | $3.92 | 1,038.69s | [deepswe](https://deepswe.datacurve.ai/artifacts/v1.1/leaderboard-live.json) |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `pass_at_1` \| 0.7323008849557522 \| score_fraction \| scored_rollout_attempts \| Higher is better \| \| `pass_at_4` \| 0.8230088495575221 \| score_fraction \| tasks_with_one_passing_rollout \| Higher is better \| \| `cost_per_attempt` \| 3.9237237909292038 \| USD \| scored_attempt \| Lower is better \| \| `duration` \| 1038.6858407079646 \| seconds \| scored_attempt \| Lower is better \| ### Configuration harness mini-swe-agent Reasoning effort high sourceConfiguration mini_swe_agent_gpt_6_astra_high ### Reported statistics Uncertainty Reported interval: 69.81–76.65% interval [0.6980659243770221,0.7665358455344823] intervalMethod 95% run-to-run: SE across repeated whole-benchmark passes (1.96 * std(runs)/sqrt(R)) runCount 4 sampleSize 452 taskCount 113 ### Source [Original source](https://deepswe.datacurve.ai/artifacts/v1.1/leaderboard-live.json) Source deepswe Original model label gpt-6-astra Observed 2026-10-22T03:50:05.889Z Evidence SHA-256 `a7c15d66288fd249c020b9931c017b92d1a3b90e480b3ff34974b752bd030019` Source locator `/rows/3` |  |  |  |  |  |  |  |
| [AutomationBench](https://benchize.com/benchmarks/zapier%3Aautomationbench.md) | 37.14% | — | 7 / 113 | $1.44 | — | [zapier](https://framerusercontent.com/sites/4WTSl4BNjd1q9QFEFibC6h/lOvLnhItj8S8-1vyjH_s33Jn-WmUXPmiJ5Diajl0Q38.CMutaHYl.mjs) |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `score` \| 37.14 \| percent \| reported \| Higher is better \| \| `cost` \| 1.44 \| USD \| per_task \| Lower is better \| ### Configuration benchmarkVersion 1.0.6 costFootnote false Fallback policy Not reported Reasoning effort high Model configuration GPT-6 Astra (High) thinkingEnabled true ### Reported statistics rank 7 ### Source [Original source](https://framerusercontent.com/sites/4WTSl4BNjd1q9QFEFibC6h/lOvLnhItj8S8-1vyjH_s33Jn-WmUXPmiJ5Diajl0Q38.CMutaHYl.mjs) Source zapier Original model label GPT-6 Astra (High) Observed 2026-10-21T16:34:56.783Z Evidence SHA-256 `43978e38f018789fa59c11de5328abe9e5905bfa9a5a4d601fa3aa4d9c153b0d` Source locator `/rows/6` |  |  |  |  |  |  |  |
| [FrontierCode 1.1 main](https://benchize.com/benchmarks/frontiercode%3Av1_1%3Amain.md) | 50.94% | — | 11 / 125 | $3.01 | — | [frontiercode](https://cognition.com/data/frontiercode-leaderboard/data.json) |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `score` \| 0.5094 \| score_fraction \| blocking_criteria_adjusted \| Higher is better \| \| `correct` \| 0.562 \| score_fraction \| raw_rubric \| Higher is better \| \| `cost` \| 3.0149 \| USD \| per_rollout \| Lower is better \| ### Configuration harness codex Reasoning effort high sourceVersion v1_1 subset main ### Reported statistics flaggedRate 0 sampleSize 100 ### Source [Original source](https://cognition.com/data/frontiercode-leaderboard/data.json) Source frontiercode Original model label GPT-6 Astra Observed 2026-10-22T05:40:03.220Z Evidence SHA-256 `ff0040778b7bdf69c705ad7474f771fb5e53cfaf5cf8573b3a4614153543d75f` Source locator `/v1_1/data/GPT-6%20Astra/high/main` |  |  |  |  |  |  |  |
| [ARC-AGI-1](https://benchize.com/benchmarks/arc-prize%3Av1_Semi_Private.md) | 98.5% | — | 1 / 245 | 0.28usd per task | — | [arc prize · v1](https://arcprize.org/media/data/leaderboard/v1.json) |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `score` \| 0.985 \| score_fraction \| v1_Semi_Private \| Higher is better \| \| `cost_per_task` \| 0.2837032 \| usd_per_task \| v1_Semi_Private \| Lower is better \| ### Configuration evaluationSplit v1_Semi_Private harness Not reported modelGroup openai-gpt-6-astra modelType CoT Reasoning effort high sourceModelVersion openai-gpt-6-astra-high thinkingEnabled true ### Source [Original source](https://arcprize.org/media/data/leaderboard/v1.json) Source arc prize Original model label GPT-6 Astra (High) Source version v1 Observed 2026-10-22T05:20:04.066Z Evidence SHA-256 `2bf68d22dafba37047d07ad833ebc0e8199c9aa2f9eeaed50c75c1b704462836` Source locator `/evaluations/226` |  |  |  |  |  |  |  |
| [ARC-AGI-2](https://benchize.com/benchmarks/arc-prize%3Av2_Semi_Private.md) | 92.08% | — | 7 / 243 | 0.67usd per task | — | [arc prize · v2](https://arcprize.org/media/data/leaderboard/v2.json) |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `score` \| 0.9208333333333333 \| score_fraction \| v2_Semi_Private \| Higher is better \| \| `cost_per_task` \| 0.6680793333333332 \| usd_per_task \| v2_Semi_Private \| Lower is better \| ### Configuration evaluationSplit v2_Semi_Private harness Not reported modelGroup openai-gpt-6-astra modelType CoT Reasoning effort high sourceModelVersion openai-gpt-6-astra-high thinkingEnabled true ### Source [Original source](https://arcprize.org/media/data/leaderboard/v2.json) Source arc prize Original model label GPT-6 Astra (High) Source version v2 Observed 2026-10-22T05:20:04.306Z Evidence SHA-256 `4351eb696e78113b0459e9bb5f0258318b905efef34f473dcea51ce641b0d94d` Source locator `/evaluations/227` |  |  |  |  |  |  |  |
| [ARC-AGI-3](https://benchize.com/benchmarks/arc-prize%3Av3_Semi_Private.md) | 54.82% | — | 3 / 65 | 40,704.72usd | — | [arc prize · v3](https://arcprize.org/media/data/leaderboard/v3.json) |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `score` \| 0.5481905163612231 \| score_fraction \| v3_Semi_Private \| Higher is better \| \| `total_cost` \| 40704.722830000006 \| usd \| v3_Semi_Private \| Lower is better \| ### Configuration evaluationSplit v3_Semi_Private harness standard modelGroup openai-gpt-6-astra modelType CoT Reasoning effort high sourceModelVersion openai-gpt-6-astra-high thinkingEnabled true ### Source [Original source](https://arcprize.org/media/data/leaderboard/v3.json) Source arc prize Original model label GPT-6 Astra (High) Source version v3 Observed 2026-10-22T05:20:04.578Z Evidence SHA-256 `3ce335487296bb2dcf4ba9546c5a2a02b1907db5b08887cd9ff4429fed824916` Source locator `/evaluations/29` |  |  |  |  |  |  |  |

---
Cite as: GPT-6 AstraHigh — Benchize, https://benchize.com/models/gpt-6-astra-high (retrieved 2026-10-22)