Unverified50% confidenceFactExact time
Exa officially claims that version 3.0 leads Perplexity, Brave and Parallel on 8 benchmarks.
1
Sources
50%
Confidence
Short-term (~14 days)
Relevance
10/20/2026
First Seen
Valid until: 11/3/2026
Sources
Related Entities
Related Claims
UnverifiedThe test was led by security firm Irregular, which had previously participated in similar tests disclosed by OpenAI, Anthropic, and Meta.61% similarUnverifiedAnthropic states that in its most comprehensive alignment testing, Opus 5.5 is the best-performing model it has tested to date.58% similarUnverifiedThe difference between Evals and traditional software integration testing lies in the probabilistic nature of LLM outputs, which requires replacing exact string assertions with a golden question set plus expected-behavior judgments.57% similarUnverifiedNext steps include adding frontier API baselines, validating the curves on a larger held-out set, and applying the findings to the next-generation ORCH model; the codebase has been open-sourced and all 7 tests pass.57% similarUnverifiedEmber-1, developed by Fairworks Research through post-training on K3, achieved the same success rate as K3 in online A/B testing of coding traffic while using 71% fewer reasoning tokens and 39% fewer total tokens.57% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1074400API
curl https://kongchang.com/api/v1/knowledge/claims/1074400MCP
get_claim(id=1074400)