Unverified50% confidenceBenchmarkExact time
On the command-line coding test, Terra's highest level raised its score from 21.5% in the previous generation to 50.3%.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
11/1/2026
First Seen
Valid until: 1/30/2027
Sources
Related Entities
Related Claims
UnverifiedOn the Terminal Bench 4.0 benchmark, GPT-6.1 Terra's score rose from 21.5% to 50.3%.69% similarUnverifiedIn Paul Well-Hurren's bug-fixing test, Terra's highest level scored 40.7, versus 32 for the previous generation's highest level; however, Terra's ultra-high level scored only 11.3 and its high level 9.7, lower than the previous generation at the same levels.68% similarUnverifiedIn Artificial Analysis's AA-Omniscience benchmark (which penalizes incorrect answers), the top score rose from 4.8 at launch to 46 currently.67% similarUnverifiedThe Program Bench score rose from 20.3 to 29.1, an improvement of 8.8 points, using the 'Almost' criterion (passing 5 of 9 tests counts as a pass).67% similarUnverifiedOn Terminal Bench, Fable 5.1's score is approximately 20 points higher than Grok 4.7's, representing close to a 50% improvement.67% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1121981API
curl https://kongchang.com/api/v1/knowledge/claims/1121981MCP
get_claim(id=1121981)