Anthropic
21 ranked models in a catalog of 28 text models. Claude Fable 5.1 leads this lab at #1 overall.
Ranked models
Global ranks · current snapshot| Rank | Model | SI Score | coding | math | reasoning | preference | Confidence | Blended price | Context | Released |
|---|---|---|---|---|---|---|---|---|---|---|
| #1 | Claude Fable 5.1 | 80.2 | 69.8 | 91.5 | 87.1 | 82.5 | 100% | $20.00 | 1M | Sep 1, 2026 |
| #2 | Claude Opus 5.5 | 78.4 | 72.5 | 91.5 | 91.4 | 82.8 | 100% | $8.00 | 1M | Sep 22, 2026 |
| #4 | Claude Fable 5 | 76.8 | 70.2 | 91.0 | 89.4 | 81.1 | 100% | $20.00 | 1M | Jun 9, 2026 |
| #5 | Claude Opus 5 | 74.9 | 69.7 | 86.9 | 84.3 | 82.0 | 100% | $10.00 | 1M | Jul 24, 2026 |
| #8 | Claude Opus 4.7 | 71.2 | 66.5 | 78.0 | 74.5 | 80.5 | 100% | $10.00 | 1M | Apr 16, 2026 |
| #10 | Claude Opus 4.6 | 70.8 | 62.4 | 74.1 | 78.8 | 81.6 | 100% | $10.00 | 1M | Feb 5, 2026 |
| #15 | Claude Sonnet 5.5 | 69.4 | 60.0 | 90.5 | 88.6 | 79.5 | 93% | $4.00 | 1M | Sep 28, 2026 |
| #20 | Claude Sonnet 4.6 | 68.3 | 60.6 | 81.9 | 80.9 | 78.4 | 100% | $6.00 | 1M | Feb 17, 2026 |
| #25 | Claude Sonnet 5 | 67.7 | 61.3 | 76.1 | 87.3 | 77.0 | 93% | $4.00 | 1M | Jun 30, 2026 |
| #26 | Claude Opus 4.8 | 67.5 | 59.6 | 82.9 | 73.7 | 78.1 | 100% | $10.00 | 1M | May 28, 2026 |
| #50 | Claude Opus 4.5 | 62.6 | 57.1 | 65.6 | 63.6 | 77.6 | 100% | $10.00 | 200K | Nov 1, 2025 |
| #62 | Claude Haiku 4.5 | 60.5 | — | 78.2 | 65.8 | 72.7 | 64% | $2.00 | 200K | Oct 15, 2025 |
| #64 | Claude Haiku 5.5 | 59.4 | 63.7 | 79.2 | 66.4 | — | 68% | $0.20 | 1M | Oct 7, 2026 |
| #87 | Claude Opus 4 | 55.5 | 70.6 | 75.2 | 22.9 | 69.6 | 100% | $30.00 | 200K | May 22, 2025 |
| #94 | Claude Sonnet 4.5 | 53.9 | 71.3 | 44.8 | 31.4 | 76.4 | 80% | $6.00 | 200K | Sep 29, 2025 |
| #101 | Claude Sonnet 3.5 v2 | 51.9 | 50.9 | 40.8 | 55.3 | 62.0 | 100% | — | 200K | Oct 22, 2024 |
| #102 | Claude Opus 4.1 | 51.5 | 73.3 | 17.6 | 30.5 | 74.9 | 80% | — | 200K | Aug 5, 2025 |
| #110 | Claude Sonnet 3.7 | 48.9 | 60.7 | 70.9 | 14.4 | 62.2 | 100% | — | 200K | Feb 19, 2025 |
| #115 | Claude Sonnet 4 | 46.6 | 45.1 | 74.8 | 21.9 | 66.7 | 100% | $6.00 | 200K | May 22, 2025 |
| #125 | Claude Haiku 3.5 | 40.8 | 34.3 | 32.3 | 38.1 | 56.9 | 100% | $1.60 | 200K | Oct 22, 2024 |
| #126 | Claude Haiku 3 | 39.9 | — | 10.5 | 36.3 | 49.3 | 93% | — | 200K | Mar 13, 2024 |
Pillars use a 0–100 scale. * marks an imputed neutral prior where the model has no scored results in that pillar; it is not a measured benchmark result. Prices are USD per 1M tokens; see the blend and price comparison.
Scores and releases
Provisional models
7 awaiting rank-eligible evidence- Claude Mythos 5
Evidence completeness 5.0%; at least 50% required
- Claude Opus 4.5 (latest)
Results cover 1 pillar; at least 2 required
- Claude Haiku 4.5 (latest)
Evidence completeness 33.0%; at least 50% required
- Claude Sonnet 4.5 (latest)
Evidence completeness 43.0%; at least 50% required
- Claude Opus 4.1 (latest)
Results cover 1 pillar; at least 2 required
- Claude Opus 4 (latest)
No usable open benchmark results found
- Claude Sonnet 4 (latest)
Results cover 1 pillar; at least 2 required