# GPT-6 AstraMedium

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

OpenAI

Canonical URL: https://benchize.com/models/gpt-6-astra-medium
Last updated: Thu, 22 Oct 2026 06:30:09 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 |  |
| --- | --- | --- | --- | --- | --- | --- | --- |
| [FrontierCode 1.1 main](https://benchize.com/benchmarks/frontiercode%3Av1_1%3Amain.md) | 48.83% | — | 22 / 125 | $2.43 | — | [frontiercode](https://cognition.com/data/frontiercode-leaderboard/data.json) |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `score` \| 0.4883 \| score_fraction \| blocking_criteria_adjusted \| Higher is better \| \| `correct` \| 0.536 \| score_fraction \| raw_rubric \| Higher is better \| \| `cost` \| 2.4259 \| USD \| per_rollout \| Lower is better \| ### Configuration harness codex Reasoning effort medium 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/medium/main` |  |  |  |  |  |  |  |
| [FrontierMath Tier 4 v2 (private)](https://benchize.com/benchmarks/epoch-ai%3Afrontiermath_tier_4_v2.md) | 97.56% | SE ±2.44 pp | 5 / 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` \| 0.975609756097561 \| score_fraction \| Best score (across scorers) \| Higher is better \| \| `mean_score` \| 0.975609756097561 \| score_fraction \| mean_score \| Higher is better \| ### Configuration benchmarkVersion FrontierMath-Tier-4-v2-Private reasoningBudget Not reported Reasoning effort medium sourceModelVersion gpt-6-astra_medium ### Reported statistics Uncertainty Standard error: ±2.44 pp standardError 0.024390243902439022 ### 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/20` |  |  |  |  |  |  |  |
| [DeepSWE](https://benchize.com/benchmarks/deepswe%3Av1.1.md) | 72.79% | Range 70.2–75.38% | 8 / 70 | $3.08 | 883.94s | [deepswe](https://deepswe.datacurve.ai/artifacts/v1.1/leaderboard-live.json) |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `pass_at_1` \| 0.7278761061946902 \| 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` \| 3.0754742975663714 \| USD \| scored_attempt \| Lower is better \| \| `duration` \| 883.9358407079646 \| seconds \| scored_attempt \| Lower is better \| ### Configuration harness mini-swe-agent Reasoning effort medium sourceConfiguration mini_swe_agent_gpt_6_astra_medium ### Reported statistics Uncertainty Reported interval: 70.2–75.38% interval [0.7019796153763715,0.7537725970130089] 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/7` |  |  |  |  |  |  |  |
| [AutomationBench](https://benchize.com/benchmarks/zapier%3Aautomationbench.md) | 34.09% | — | 10 / 113 | $1.27 | — | [zapier](https://framerusercontent.com/sites/4WTSl4BNjd1q9QFEFibC6h/lOvLnhItj8S8-1vyjH_s33Jn-WmUXPmiJ5Diajl0Q38.CMutaHYl.mjs) |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `score` \| 34.09 \| percent \| reported \| Higher is better \| \| `cost` \| 1.27 \| USD \| per_task \| Lower is better \| ### Configuration benchmarkVersion 1.0.6 costFootnote false Fallback policy Not reported Reasoning effort medium Model configuration GPT-6 Astra (Medium) thinkingEnabled true ### Reported statistics rank 10 ### Source [Original source](https://framerusercontent.com/sites/4WTSl4BNjd1q9QFEFibC6h/lOvLnhItj8S8-1vyjH_s33Jn-WmUXPmiJ5Diajl0Q38.CMutaHYl.mjs) Source zapier Original model label GPT-6 Astra (Medium) Observed 2026-10-21T16:34:56.783Z Evidence SHA-256 `43978e38f018789fa59c11de5328abe9e5905bfa9a5a4d601fa3aa4d9c153b0d` Source locator `/rows/9` |  |  |  |  |  |  |  |
| [ARC-AGI-2](https://benchize.com/benchmarks/arc-prize%3Av2_Semi_Private.md) | 92.08% | — | 7 / 243 | 0.48usd 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` \| 0.4800926666666667 \| 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 medium sourceModelVersion openai-gpt-6-astra-medium thinkingEnabled true ### Source [Original source](https://arcprize.org/media/data/leaderboard/v2.json) Source arc prize Original model label GPT-6 Astra (Medium) Source version v2 Observed 2026-10-22T05:20:04.306Z Evidence SHA-256 `4351eb696e78113b0459e9bb5f0258318b905efef34f473dcea51ce641b0d94d` Source locator `/evaluations/228` |  |  |  |  |  |  |  |
| [Terminal-Bench 4](https://benchize.com/benchmarks/harbor%3Aterminal-bench-4.md) | 54.24% | 95% CI 51.58–56.9% | 5 / 17 | — | — | [harbor](https://www.tbench.ai/) |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `accuracy` \| 54.24 \| percent \| reported \| Higher is better \| \| `pass_at_2` \| 0.603 \| score_fraction \| reported \| Higher is better \| \| `pass_at_3` \| 0.6318 \| score_fraction \| reported \| Higher is better \| \| `pass_at_4` \| 0.6515 \| score_fraction \| reported \| Higher is better \| \| `pass_at_5` \| 0.6667 \| score_fraction \| reported \| Higher is better \| ### Configuration agent Codex Reasoning effort medium sourceSubmission f3c3d5a6-6424-4acb-bfcc-615c3f79f6dd ### Reported statistics Uncertainty 95% confidence interval: 51.58–56.9% confidenceHalfWidth95 2.66 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/9` |  |  |  |  |  |  |  |
| [ARC-AGI-3](https://benchize.com/benchmarks/arc-prize%3Av3_Semi_Private.md) | 38.59% | — | 8 / 65 | 48,090.35usd | — | [arc prize · v3](https://arcprize.org/media/data/leaderboard/v3.json) |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `score` \| 0.3858757449733784 \| score_fraction \| v3_Semi_Private \| Higher is better \| \| `total_cost` \| 48090.34598999999 \| usd \| v3_Semi_Private \| Lower is better \| ### Configuration evaluationSplit v3_Semi_Private harness standard modelGroup openai-gpt-6-astra modelType CoT Reasoning effort medium sourceModelVersion openai-gpt-6-astra-medium thinkingEnabled true ### Source [Original source](https://arcprize.org/media/data/leaderboard/v3.json) Source arc prize Original model label GPT-6 Astra (Medium) Source version v3 Observed 2026-10-22T05:20:04.578Z Evidence SHA-256 `3ce335487296bb2dcf4ba9546c5a2a02b1907db5b08887cd9ff4429fed824916` Source locator `/evaluations/30` |  |  |  |  |  |  |  |
| [FrontierCode 1.1 extended](https://benchize.com/benchmarks/frontiercode%3Av1_1%3Aextended.md) | 60.25% | — | 29 / 125 | $2.1 | — | [frontiercode](https://cognition.com/data/frontiercode-leaderboard/data.json) |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `score` \| 0.6025 \| score_fraction \| blocking_criteria_adjusted \| Higher is better \| \| `correct` \| 0.6573 \| score_fraction \| raw_rubric \| Higher is better \| \| `cost` \| 2.102 \| USD \| per_rollout \| Lower is better \| ### Configuration harness codex Reasoning effort medium 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/medium/extended` |  |  |  |  |  |  |  |
| [ToneBench Scriptwriting](https://benchize.com/benchmarks/tonebench%3Ascripts.md) | 1,979.4 | Range 1,946.1–2,013 | 36 / 126 | — | — | [tonebench](https://benchmark.towardsai.com/data/leaderboard.json) |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `elo` \| 1979.4 \| rating \| reported \| Higher is better \| \| `overall` \| 80.67 \| percent \| reported \| Higher is better \| \| `anti_slop_numbers` \| 80 \| source_score \| rubric \| Higher is better \| \| `continuity_emotion` \| 80 \| source_score \| rubric \| Higher is better \| \| `cue_quality` \| 80.69 \| source_score \| rubric \| Higher is better \| \| `hook_strength` \| 82.46 \| source_score \| rubric \| Higher is better \| \| `length_adherence` \| 83.22 \| source_score \| rubric \| Higher is better \| \| `slop` \| 88.5 \| source_score \| rubric \| Higher is better \| \| `slop_score` \| 97.01 \| source_score \| rubric \| Higher is better \| \| `substance_accuracy` \| 86.02 \| source_score \| rubric \| Higher is better \| \| `tone_voice` \| 76.98 \| source_score \| rubric \| Higher is better \| \| `writing_quality` \| 78.39 \| source_score \| rubric \| Higher is better \| \| `youtube_best_practices` \| 76.61 \| source_score \| rubric \| Higher is better \| ### Configuration provider codex Reasoning effort unknown scoreScope 10_task_current sourceConfiguration gpt-6-astra thinkingEnabled true ### Reported statistics Uncertainty Reported interval: 1,946.1–2,013 interval [1946.1,2013] rank 36 runCount 5 ### Source [Original source](https://benchmark.towardsai.com/data/leaderboard.json) Source tonebench Original model label GPT-6 Astra (medium) Evaluated 2026-10-11 Observed 2026-10-22T04:40:14.102Z Evidence SHA-256 `4264338337f418d3ded7ace250cad64fe12cacbd4b79a312bfaf0ae9268a01c3` Source locator `/rows/35` |  |  |  |  |  |  |  |
| [ARC-AGI-1](https://benchize.com/benchmarks/arc-prize%3Av1_Semi_Private.md) | 97.5% | — | 12 / 245 | 0.24usd per task | — | [arc prize · v1](https://arcprize.org/media/data/leaderboard/v1.json) |  |
| ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `score` \| 0.975 \| score_fraction \| v1_Semi_Private \| Higher is better \| \| `cost_per_task` \| 0.2398342000000002 \| 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 medium sourceModelVersion openai-gpt-6-astra-medium thinkingEnabled true ### Source [Original source](https://arcprize.org/media/data/leaderboard/v1.json) Source arc prize Original model label GPT-6 Astra (Medium) Source version v1 Observed 2026-10-22T05:20:04.066Z Evidence SHA-256 `2bf68d22dafba37047d07ad833ebc0e8199c9aa2f9eeaed50c75c1b704462836` Source locator `/evaluations/227` |  |  |  |  |  |  |  |

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