Cloud BenchmarksLocal Benchmarks

☁️ Cloud Benchmarks ☁️

I run cron jobs to periodically test the token generation speed of different cloud LLM providers. The chart helps visualize the distributions of different speeds, as they can vary somewhat depending on the loads. For readability not all models are shown, but you can see the full results in the table below.

Every provider and model now has a dedicated landing page with narrative insights, SEO-friendly metadata, and structured data for search engines. Click any provider or model in the table to explore performance in depth.

I am working daily to add more providers and models, looking anywhere that does not require purchasing dedicated endpoints for hosting (why some models may appear to be missing). If you have any more suggestions let me know on GitHub!! 😊

Fastest Models Right Now (updated <24h)

#ModelProviderSpeed
1llama-3.1-8bgroq281 tok/s
2qwen-3-32bcerebras254 tok/s
3qwen-3-32bgroq240 tok/s
4llama-3.1-8bcerebras224 tok/s
5llama-3.3-70bcerebras216 tok/s

📊 Speed Distribution 📊

📚 Full Results 📚

Showing 86 of 86 modelsFlagged statuses: likely_deprecated, deprecated, failing, stale, never_succeeded, disabled
Status
groqllama-3.1-8bActive1h ago281.0095447130.00
cerebrasqwen-3-32bActive1h ago254.004417360.00
groqqwen-3-32bActive1h ago240.0046391160.00
cerebrasllama-3.1-8bActive1h ago224.006365540.00
cerebrasllama-3.3-70bActive1h ago216.0014338370.00
cerebrasgpt-oss-120bActive1h ago213.004346750.00
groqllama-3.3-70bActive1h ago206.0079280120.00
groqllama-4-scoutActive1h ago195.0023316250.00
groqllama-4-maverickActive1h ago165.0019310510.00
cerebrasqwen-3-235b-instructActive4d ago163.002264790.00
togetherllama-3.1-8bActive1h ago155.002230580.00
groqkimi-k2Active1h ago140.0021203250.00
bedrocknova-microActive17m ago128.0069154260.00
bedrockllama-4-maverickActive17m ago109.0028149250.00
togethermistral-7bActive1h ago109.003166410.00
openaio3 MiniNever Succeeded(Medium)1h ago104.00201510.00
bedrocknova-liteActive17m ago102.0042135290.00
bedrockllama-4-scoutActive17m ago102.001140340.00
bedrockllama-3.3-70bActive17m ago101.0015137230.00
togetherqwen-2.5-7bActive1h ago100.003146270.00
bedrocknova-proActive17m ago82.7010124410.00
openaigpt-3.5-turboActive1h ago77.5012129460.00
openaigpt-4.1-nanoActive1h ago77.1013138400.00
googleActive1h ago74.3034105580.00
togetherllama-3.2-3bActive1h ago73.2051451150.00
togetherllama-3.1-70bActive1h ago72.604147480.00
googleclaude-3-haikuActive20d ago67.602779530.00
openaigpt-4oActive1h ago64.0071411410.00
fireworksmixtral-8x22bActive1h ago61.503783570.00
googlegemini-2.5-flashNever Succeeded(Medium)1h ago61.507781180.00
googlegemini-2.0-flashActive4d ago61.40888590.00
googlegemini-2.0-flash-liteActive4d ago61.101181610.00
togetherllama-3.3-70bActive1h ago52.0021361470.00
anthropicclaude-haiku-4.5Active1h ago51.501580620.00
togetherqwen-2.5-72bActive1d ago50.90371490.00
deepinframixtral-8x22bStale(Medium)1h ago50.002680300.00
openaio4 MiniNever Succeeded(Medium)1h ago49.6015750.00
togethermixtral-8x7bActive1h ago49.206111330.00
bedrockllama-3.2-90bActive17m ago47.402951340.00
openaigpt-4.1-miniActive1h ago47.001897480.00
deepinfrallama-3-8bStale(Medium)1h ago45.60771300.00
bedrockclaude-haiku-4.5Active17m ago45.40865890.00
deepinframistral-7bStale(Medium)1h ago45.10389540.00
bedrockmistral-largeActive17m ago44.00847240.00
fireworksllama-3.3-70bActive1h ago41.906781250.00
deepinfradevstral-smallNever Succeeded(Medium)1h ago41.80384510.00
googlegemini-2.5-proNever Succeeded(Medium)1h ago41.4015631570.00
openaigpt-4o-miniActive1h ago39.90886420.00
openaiGPT-5.1-codex-maxActive1h ago39.7011041500.00
deepinfrallama-3.2-90bStale(Medium)1h ago38.60193860.00
togetherdeepseek-r1Active1h ago38.001681770.00
deepinfrallama-3.1-8bStale(Medium)1h ago37.5011021250.00
googleclaude-3-5-sonnetActive20d ago34.702446720.00
openaigpt-4.1Active1h ago34.50767450.00
openaigpt-4-turboActive1h ago33.90251560.00
deepinfrallama-2-70bStale(Medium)1h ago33.20844390.00
bedrockclaude-3-7-sonnetActive17m ago32.90142790.00
deepinfrallama-3-70bStale(Medium)1h ago32.80244530.00
bedrockclaude-3-5-sonnetActive17m ago32.60742570.00
deepinfraqwen-2.5-72bStale(Medium)1h ago31.90247870.00
deepinfraQwen 2.5 Coder 32BNever Succeeded(Medium)1h ago31.701674650.00
bedrockclaude-3-5-haikuActive17m ago29.001381260.00
togetherdeepseek-v3Active15d ago28.801671400.00
openaiGPT-5.1Active1h ago28.003561090.00
openaiGPT-5.2Active1h ago27.60944990.00
deepinfrallama-3.2-3bStale(Medium)1h ago27.20190640.00
deepinfrallama-3.2-1bStale(Medium)1h ago26.70483370.00
openaigpt-4Active1h ago26.20252600.00
openaiGPT-5.1-codexActive1h ago25.80449890.00
deepinfrallama-3.3-70bNever Succeeded(Medium)1h ago23.50159820.00
deepinfraqwen-3-235bNever Succeeded(Medium)1h ago22.00143530.00
bedrockclaude-sonnet-4.5Active17m ago21.601291800.00
togetherllama-3.1-405bActive2d ago21.401301420.00
deepinfrallama-3.1-70bStale(Medium)1h ago21.20339530.00
openaiGPT-5.1-codex-miniActive1h ago21.001451030.00
anthropicclaude-4-sonnetActive1h ago20.408321570.00
anthropicclaude-opus-4.5Active1h ago19.7011301720.00
bedrockclaude-3-opusActive17m ago19.00422850.00
anthropicClaude Opus 4.1Active1h ago19.005251320.00
anthropicclaude-4-opusActive1h ago18.104241190.00
bedrockclaude-opus-4.5Active17m ago18.001232240.00
deepinfrallama-3.1-405bStale(Medium)1h ago17.201312920.00
deepinfrallama-3.2-11bStale(Medium)1h ago14.201681990.00
openaiActive1h ago11.906182100.00
openaio1-proLikely Deprecated(Medium)4h ago9.86119100.00
openaiGPT-5.2-proActive1h ago4.311125650.00
Lifecycle snapshot
Loading status summary…

📈 Time Series 📈

llama-3.3-70b

llama-3.1-8b

claude-3-5-sonnet

claude-haiku-4.5

claude-opus-4.5

llama-3.1-405b

llama-3.1-70b

llama-3.2-3b

llama-3.2-90b

llama-4-maverick

llama-4-scout

mistral-7b

mixtral-8x22b

qwen-2.5-72b

qwen-3-32b

undefined

Claude Opus 4.1

claude-3-5-haiku

claude-3-7-sonnet

claude-3-opus

claude-4-opus

claude-4-sonnet

claude-sonnet-4.5

deepseek-r1

devstral-small

gemini-2.0-flash

gemini-2.0-flash-lite

gemini-2.5-flash

gemini-2.5-pro

gpt-3.5-turbo

gpt-4

gpt-4-turbo

gpt-4.1

gpt-4.1-mini

gpt-4.1-nano

gpt-4o

gpt-4o-mini

GPT-5.1

GPT-5.1-codex

GPT-5.1-codex-max

GPT-5.1-codex-mini

GPT-5.2

GPT-5.2-pro

gpt-oss-120b

kimi-k2

llama-2-70b

llama-3-70b

llama-3-8b

llama-3.2-11b

llama-3.2-1b

mistral-large

mixtral-8x7b

nova-lite

nova-micro

nova-pro

o1-pro

o3 Mini

o4 Mini

Qwen 2.5 Coder 32B

qwen-2.5-7b

qwen-3-235b

qwen-3-235b-instruct