OpenAI’s GPT-6.2 Astra scores 174 on Epoch Capabilities Index, tying Claude Fable 5.2 OpenAI's GPT-6.2 Astra scored 174 on the Epoch Capabilities Index, the composite benchmark run by Epoch AI, tying Anthropic's Claude Fable 5.2, which posted the same 174 shortly after its November 9, 2026 launch. Epoch AI published the GPT-6.2 Astra evaluation on November 17, 2026, placing it above the pre-Astra trend line for reasoning models and 5 to 7 points above the 167-169 range earlier GPT-6 Astra models recorded between September and November 2026. The 174 score sits 24 points above the GPT-5 anchor of 150 and 44 points above Claude 3.5 Sonnet's 130, and the update used third-party evaluations rather than self-reported benchmarks from either company. OpenAI’s GPT-6.2 Astra scores 174 on Epoch Capabilities Index, tying Claude Fable 5.2 Photo: Tim Witzdam / Pexels The new score beats the pre-Astra trend line for reasoning models and puts OpenAI level with Anthropic's latest release OpenAI https://cryptobriefing.com/markets/openai/ ’s GPT-6.2 Astra has posted a score of 174 on the Epoch Capabilities Index ECI , the composite AI benchmark run by Epoch AI. That puts it exactly level with Anthropic https://cryptobriefing.com/markets/anthropic/ ’s Claude Fable 5.2. The result also lands above the pre-Astra trend for reasoning models, according to Epoch AI. What the numbers show Epoch AI published the updated evaluation for GPT-6.2 Astra on November 17, 2026. That came just over a week after Anthropic launched Claude Fable 5.2 on November 9, 2026. Fable 5.2 recorded its own score of 174 shortly after release. So two companies, using different approaches and shipping on different schedules, arrived at the same number within days of each other. The bigger story may be the jump inside OpenAI’s own lineup. Earlier models in the GPT-6 Astra line scored between 167 and 169 on the ECI in assessments running from September to November 2026. Moving from that range to 174 is a gain of 5 to 7 points in a matter of weeks. AI, tech, and the markets they move—in one daily briefing. Daily. Free. Join 34,000+ readers across crypto, finance, and policy. Epoch AI frames the result against the pre-Astra trend for reasoning models. GPT-6.2 Astra sits above that line. Fable 5.2, at the same score, does too. How the Epoch Capabilities Index works The ECI rolls more than 50 benchmarks across a range of disciplines into a single composite score. The scale is anchored to familiar reference points. Claude 3.5 Sonnet is pinned at 130, and GPT-5 sits at 150. Those anchors give the 174 score some context. Both GPT-6.2 Astra and Claude Fable 5.2 now sit 24 points above the GPT-5 marker and 44 points above Claude 3.5 Sonnet. No self-reported benchmarks from OpenAI or Anthropic went into this ECI update. The scores come from third-party evaluations. Background: a crowded frontier The GPT-6 Astra line has been on a steady march through the autumn of 2026. Scores of 167 to 169 across September through early November suggested incremental improvement. GPT-6.2 Astra broke out of that band. Anthropic’s Fable 5.2 got there first, landing its 174 shortly after its November 9 launch. OpenAI’s model matched it in an evaluation published on November 17. Disclosure: This article was edited by Diego Almada Lopez. For more information on how we create and review content, see our Editorial Policy https://cryptobriefing.com/editorial-policy/ .