The Vikshy Score is a daily 0 to 100 reading of each major AI provider's competitive strength and market pull, computed by a deterministic, published method from sourced, evidence-graded developments. As of 2026-07-24, OpenAI leads at 85.3, with 8 providers ranked and more watched provisionally.
Where a provider stands today across capability, adoption, developer pull, and reliability, then tilted up or down by its 30-day momentum. Standing sets the level; momentum bends the line.
A single number on a 0 to 100 scale, with a 30-day trend and a confidence grade. Higher means stronger position and direction relative to the other providers, not a "best model" verdict.
It is deterministic: same evidence in, same score out. Every input is graded by source tier, and every number is traceable to cited evidence in the log. No opinions, no vendor payments.
Not a valuation, not investment advice, and not a claim about which model is best for your specific task. Community signal is capped and can never establish a fact on its own.
The Vikshy Score is a daily 0 to 100 reading of an AI provider's current competitive strength and market pull, computed by a deterministic, published methodology from sourced, evidence-graded developments. It is not a "best AI" verdict, a valuation, or investment advice.
Each of seven factors is scored on where a provider stands today, then tilted by its 30-day momentum (Factor = Standing + 0.35 × (Momentum − 50), clamped 0 to 100). The Score is the weighted sum of those factors. Same evidence in, same score out, and every number traces to cited evidence.
A higher score means stronger current standing and momentum across capability, adoption, developer pull, and reliability, relative to the other providers on the Board. It is a reading of position and direction, not a claim that one model is "best" for your task.
The Board recomputes daily at 06:00 UTC. Stories and flash updates flow through the day as evidence arrives; the ranked Score itself moves once a day.
Vikshy computes it from a versioned, public evidence ledger. Every input is graded by source tier, every factor is published, and each provider's number is traceable in the evidence log, so the whole calculation is auditable.