API Status

deepinfra Provider Benchmarks

Comprehensive performance summary covering 17 models.

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

Visit deepinfra Official Website

Provider Snapshot

Models Tracked

17

Avg Tokens / Second

36.71

Avg Time to First Token (ms)

1317.06

Last Updated

Mar 8, 2026

Key Takeaways

  • 17 deepinfra models are actively benchmarked with 3806 total measurements across 3736 benchmark runs.

  • mistral-7b leads the fleet with 66.30 tokens/second, while llama-3.2-1b delivers 43.00 tok/s.

  • Performance varies by 54.2% across the deepinfra model lineup, indicating diverse optimization strategies for different use cases.

  • Avg time to first token across the fleet is 1317.06 ms, showing moderate responsiveness for interactive applications.

  • The deepinfra model fleet shows varied performance characteristics (42.7% variation coefficient), reflecting diverse model architectures.

Fastest Models

ProviderModelAvg Toks/SecMinMaxAvg TTF (ms)
deepinframistral-7b66.305.62139.00620.00
deepinfradevstral-small60.709.24138.00600.00
deepinframixtral-8x22b58.9014.4077.50400.00
deepinfrallama-3.1-8b49.104.4895.30540.00
deepinfrallama-3-8b45.609.3270.30310.00
deepinfrallama-3.2-1b43.001.92100.00620.00

All Models

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

ProviderModelAvg Toks/SecMinMaxAvg TTF (ms)
deepinfraQwen 2.5 Coder 32B31.201.0774.202660.00
deepinfradevstral-small60.709.24138.00600.00
deepinfrallama-3.3-70b18.801.0945.902520.00
deepinfraqwen-3-235b10.901.0539.305850.00
deepinfraqwen-2.5-72b32.402.6849.80680.00
deepinframistral-7b66.305.62139.00620.00
deepinframixtral-8x22b58.9014.4077.50400.00
deepinfrallama-3.2-90b34.703.7175.10870.00
deepinfrallama-3.2-11b13.701.8061.102630.00
deepinfrallama-3.2-3b42.201.0397.50790.00
deepinfrallama-3.2-1b43.001.92100.00620.00
deepinfrallama-3.1-405b24.701.3636.00990.00
deepinfrallama-3.1-70b22.201.0245.501070.00
deepinfrallama-3.1-8b49.104.4895.30540.00
deepinfrallama-3-70b34.602.3554.70680.00
deepinfrallama-3-8b45.609.3270.30310.00
deepinfrallama-2-70b35.103.2456.70560.00

Featured Models

Frequently Asked Questions

Based on recent tests, mistral-7b shows the highest average throughput among tracked deepinfra models.

This provider summary aggregates 3806 individual prompts measured across 3736 monitoring runs over the past month.