VECTOR WIREAI INTELLIGENCE
⌘K
NVDA$1,847+3.2%MSFT$512+1.1%GOOGL$199-0.4%META$728+2.7%AMD$184-1.2%TSM$212+0.6%PLTR$98+4.1%AI IDX4,821+1.9%
PKT
SEEDRefresh Models —Deals —Regulatory —Sources —

mrcr_v2_8needle_128k_average

20 models tracked

MRCR v2 (OpenAI long-context needle benchmark) — Explicit v2, 8-needle, 128K context, averaged metric; Google Gemini source.

#ModelVendorBest scoreRunsLast seen
1Gemini 3.7 FlashGoogle9712026-08-14
2GPT-5.6 TerraOpenAI93.512026-08-14
3Gemini 3.6 FlashGoogle91.812026-07-22
4Gemini 3.1 ProGoogle84.932026-08-24
5Claude Sonnet 4.6Anthropic84.922026-07-06
6Claude Opus 4.6Anthropic8412026-05-01
7GPT-5.2OpenAI83.812026-05-01
8Claude Sonnet 5Anthropic81.512026-07-22
9Grok4.5SpaceXAI81.412026-07-22
10Gemini 3.5 FlashGoogle77.322026-08-24
11Gemini 3 ProGoogle7712026-05-01
12GPT-5.6 LunaOpenAI74.812026-07-22
13Gemini 3 Flash PreviewGoogle67.212026-07-06
14Gemma 4 31BGoogle66.4262026-06-06
15Claude Opus 4.7Anthropic59.322026-07-06
16Gemma 4 26B A4BGoogle44.1262026-06-06
17Gemma 4 12BGoogle43.4262026-07-17
18Gemma 4 E4BGoogle25.4312026-07-11
19Gemma 4 E2BGoogle19.1352026-07-21
20Gemma 3 27BGoogle13.5272026-07-10

Best tracked score per model (default configuration; source-attributed and verification-tiered). Open a model for its full benchmark surface, provenance and pricing.