Compare AI models
| Benchmarks | ||
SWE-Bench VerifiedCoding — Real coding tasks pulled from open-source projects — the AI has to find and fix actual bugs. A human-checked version of the original SWE-Bench. Higher is better. | 40.6% | 49% |
BullshitBench v2Nonsense detection — Given a confidently-worded but nonsensical prompt, does the AI spot that it makes no sense and push back — instead of playing along and inventing an answer? The score is how often it clearly called out the nonsense. Higher is better. | 50% | — |
GPQA DiamondScience — Graduate-level science questions in biology, physics, and chemistry — hard enough that subject-matter PhDs score around 65%. Higher is better. | 41.6% | — |
| Overview | ||
| Company | Anthropic | Anthropic |
| Release date | Oct 22 2024 | Oct 22 2024 |
| Access | Closed | Closed |
| Model details | View model | View model |
Other comparisons
Claude 3.5 HaikuvsGPT-6.1 SolClaude 3.5 Sonnet (upgraded)vsGPT-6.1 SolClaude 3.5 HaikuvsGemini 4 ArgonClaude 3.5 Sonnet (upgraded)vsGemini 4 ArgonClaude 3.5 HaikuvsMuse Spark 1.3Claude 3.5 Sonnet (upgraded)vsMuse Spark 1.3Claude 3.5 HaikuvsGrok 4.7Claude 3.5 Sonnet (upgraded)vsGrok 4.7Claude 3.5 HaikuvsDeepSeek-V4.1-FlashClaude 3.5 Sonnet (upgraded)vsDeepSeek-V4.1-FlashClaude 3.5 HaikuvsMistral Large 4Claude 3.5 Sonnet (upgraded)vsMistral Large 4Frequently asked questions
Claude 3.5 Sonnet (upgraded) leads Claude 3.5 Haiku on 1 of the 1 benchmark they both report (SWE-Bench Verified). Both models were released on the same day — Oct 22 2024.
Published specifications for these two models are limited — see each model page for the latest details.
On SWE-Bench Verified, Claude 3.5 Sonnet (upgraded) leads at 49% vs Claude 3.5 Haiku at 40.6%.