Search · KernelIndex

Supported 9sorted byRecommendedNewest

Has source9 License known9Installable0

Claims support; no run on this exact workload yet.

Implementation

Latency

Trust

Observed

—

claude-opus-4-1-20250805 / cuda8688f2claude-opus-4-1-20250805

—

— · Apache-2.0 · source

—

Declared support only; no measurement for this workload.

source mirroredApache-2.0no install recipeView source →

—

claude-opus-4-1-20250805_triton_002913claude-opus-4-1-20250805

—

— · Apache-2.0 · source

—

Declared support only; no measurement for this workload.

source mirroredApache-2.0no install recipeView source →

—

gemini-2.5-pro / cuda0ade42gemini-2.5-pro

—

— · Apache-2.0 · source

—

Declared support only; no measurement for this workload.

source mirroredApache-2.0no install recipeView source →

—

gemini-2.5-pro / triton2a8f55gemini-2.5-pro

—

— · Apache-2.0 · source

—

Declared support only; no measurement for this workload.

source mirroredApache-2.0no install recipeView source →

—

gpt-5-2025-08-07 / cudad68ec9gpt-5-2025-08-07

—

— · Apache-2.0 · source

—

Declared support only; no measurement for this workload.

source mirroredApache-2.0no install recipeView source →

—

gpt-5-2025-08-07 / triton7230f5gpt-5-2025-08-07

—

— · Apache-2.0 · source

—

Declared support only; no measurement for this workload.

source mirroredApache-2.0no install recipeView source →

—

gpt-o3 / cuda0002fbgpt-o3

—

— · Apache-2.0 · source

—

Declared support only; no measurement for this workload.

source mirroredApache-2.0no install recipeView source →

—

gpt-o3 / triton1d8355gpt-o3

—

— · Apache-2.0 · source

—

Declared support only; no measurement for this workload.

source mirroredApache-2.0no install recipeView source →

—

flashinfer_wrapper_d86b24bdbaselineFlashInfer-Bench baselines

—

— · Apache-2.0 · source

—

Declared support only; no measurement for this workload.

source mirroredApache-2.0no install recipeView source →

Top k sampling from probs v129280Operation in the sampling familyTop k sampling from probs v151936Operation in the sampling familyTop k sampling from probs v152064Operation in the sampling familyTop k sampling from probs v163840Operation in the sampling familyTop k sampling from probs v200064Operation in the sampling familyTop k top p sampling from probs v128256Operation in the sampling family

Missing a kernel here? Evidence submissions open with the contribution beta.

This search as an API call

curlkiJSON

Supported 9sorted byRecommendedNewest

Has source9 License known9Installable0

Claims support; no run on this exact workload yet.

Implementation

Latency

Trust

Observed

—

claude-opus-4-1-20250805 / cuda8688f2claude-opus-4-1-20250805

—

— · Apache-2.0 · source

—

Declared support only; no measurement for this workload.

source mirroredApache-2.0no install recipeView source →

—

claude-opus-4-1-20250805_triton_002913claude-opus-4-1-20250805

—

— · Apache-2.0 · source

—

Declared support only; no measurement for this workload.

source mirroredApache-2.0no install recipeView source →

—

gemini-2.5-pro / cuda0ade42gemini-2.5-pro

—

— · Apache-2.0 · source

—

Declared support only; no measurement for this workload.

source mirroredApache-2.0no install recipeView source →

—

gemini-2.5-pro / triton2a8f55gemini-2.5-pro

—

— · Apache-2.0 · source

—

Declared support only; no measurement for this workload.

source mirroredApache-2.0no install recipeView source →

—

gpt-5-2025-08-07 / cudad68ec9gpt-5-2025-08-07

—

— · Apache-2.0 · source

—

Declared support only; no measurement for this workload.

source mirroredApache-2.0no install recipeView source →

—

gpt-5-2025-08-07 / triton7230f5gpt-5-2025-08-07

—

— · Apache-2.0 · source

—

Declared support only; no measurement for this workload.

source mirroredApache-2.0no install recipeView source →

—

gpt-o3 / cuda0002fbgpt-o3

—

— · Apache-2.0 · source

—

Declared support only; no measurement for this workload.

source mirroredApache-2.0no install recipeView source →

—

gpt-o3 / triton1d8355gpt-o3

—

— · Apache-2.0 · source

—

Declared support only; no measurement for this workload.

source mirroredApache-2.0no install recipeView source →

—

flashinfer_wrapper_d86b24bdbaselineFlashInfer-Bench baselines

—

— · Apache-2.0 · source

—

Declared support only; no measurement for this workload.

source mirroredApache-2.0no install recipeView source →

Missing a kernel here? Evidence submissions open with the contribution beta.

This search as an API call

curlkiJSON

Top k sampling from probs v128256

Related

Top k sampling from probs v128256

Related