# MRCR v2 (8-needle) (128k average) — AI model rankings

Tests whether the AI can find specific details buried inside a very long document (around 128k tokens — roughly a long book). Higher is better.

6 tracked models have a published MRCR v2 (8-needle) (128k average) score. Higher is better. Scores are as published at each model's release.

## Ranking

| Rank | Model | Developer | Score | Released |
| --- | --- | --- | --- | --- |
| 1 | GPT-5.5 | OpenAI | 94.8% | Apr 23 2026 |
| 2 | Claude Sonnet 4.6 | Anthropic | 84.9% | Feb 17 2026 |
| 2 | Gemini 3.1 Pro | Google | 84.9% | Feb 19 2026 |
| 4 | Gemini 3.5 Flash | Google | 77.3% | May 19 2026 |
| 5 | Gemini 3.0 Flash | Google | 67.2% | Dec 17 2025 |
| 6 | Claude Opus 4.7 | Anthropic | 59.3% | Apr 16 2026 |


---

Canonical page: https://aireleasetracker.com/benchmark/mrcr-v2-128k
Full dataset: https://aireleasetracker.com/llms-full.txt · JSON: https://aireleasetracker.com/models.json
Source: AI Release Tracker (https://aireleasetracker.com). Benchmark scores are the figures published by the releasing lab at launch.