API Status

anthropic Provider Benchmarks

Comprehensive performance summary covering 7 models.

This provider hub highlights throughput and latency trends across every anthropic model monitored by LLM Benchmarks. Use it to compare hosting tiers, track regressions, and discover the fastest variants in the catalogue.

Visit anthropic Official Website

Provider Snapshot

Models Tracked

7

Avg Tokens / Second

25.64

Avg Time to First Token (ms)

1714.29

Last Updated

Jun 25, 2026

Key Takeaways

  • 7 anthropic models are actively benchmarked with 1350 total measurements across 1246 benchmark runs.

  • claude-haiku-4.5 leads the fleet with 45.20 tokens/second, while claude-4-opus delivers 15.80 tok/s.

  • Performance varies by 186.1% across the anthropic model lineup, indicating diverse optimization strategies for different use cases.

  • Avg time to first token across the fleet is 1714.29 ms, showing moderate responsiveness for interactive applications.

  • The anthropic model fleet shows varied performance characteristics (38.0% variation coefficient), reflecting diverse model architectures.

Fastest Models

ProviderModelAvg Toks/SecMinMaxAvg TTF (ms)
anthropicclaude-haiku-4.545.207.3464.90800.00
anthropicclaude-opus-4-831.705.0145.601260.00
anthropicclaude-4-sonnet26.309.8233.70850.00
anthropicclaude-sonnet-4.623.907.6930.701250.00
anthropicclaude-opus-4.522.705.0335.301440.00
anthropicclaude-4-opus15.807.9719.901640.00

All Models

Complete list of all anthropic models tracked in the benchmark system. Click any model name to view detailed performance data.

ProviderModelAvg Toks/SecMinMaxAvg TTF (ms)
anthropicclaude-fable-513.908.9316.604760.00
anthropicclaude-haiku-4.545.207.3464.90800.00
anthropicclaude-4-opus15.807.9719.901640.00
anthropicclaude-opus-4.522.705.0335.301440.00
anthropicclaude-opus-4-831.705.0145.601260.00
anthropicclaude-4-sonnet26.309.8233.70850.00
anthropicclaude-sonnet-4.623.907.6930.701250.00

Featured Models

Frequently Asked Questions

Based on recent tests, claude-haiku-4.5 shows the highest average throughput among tracked anthropic models.

This provider summary aggregates 1350 individual prompts measured across 1246 monitoring runs over the past month.