Unverified50% confidenceBenchmarkExact time
On Exploit Gym, Opus 5 solved about 190 exploits, Opus 5.5 solved 300, and Mythos/Fable 5.2 solved 341 (out of 869 in total); Gemini Carbon is extrapolated to solve all of them around 2028.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
11/11/2026
First Seen
Valid until: 2/9/2027
Sources
Related Entities
Related Claims
UnverifiedFable 5.2 beat every model on the market on the vast majority of benchmarks, losing only to Gemini 4 Carbon on the commercial workflow test and to Opus 5.5 on the knowledge work test.63% similarVerifiedGemini 4 Carbon has a vulnerability discovery rate of 88.2% and a vulnerability patching rate of 71.0%.63% similarUnverified2024年Gemini 1.5 Pro将上下文上限推至100万token63% similarUnverifiedAccording to an official Perplexity tweet, Gemini 4 Carbon's capability on agent research tasks is similar to that of Fable 5.2.61% similarUnverifiedOn most tasks, Opus 5.5's performance is comparable to Fable 5.1.61% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1157853API
curl https://kongchang.com/api/v1/knowledge/claims/1157853MCP
get_claim(id=1157853)