Preferences and analytics

BenchSift keeps your language, theme, and comparison preferences in your browser. Audience measurements also help improve the site.

Back

InclusionAI

Ling 3.0 Flash

24.9 Intelligence Index2026-08-04ReasoningOpen weightsHugging Face
Artificial Analysis Indices
Default
Intelligence24.9
Coding50.6
Agentic21.0
Standard Benchmarks
GPQA Diamond85.5%
HLE23.7%
SciCode42.0%
AA-LCR73.0%
Terminal-Bench 2.155.4%
τ-Bench Banking27.2%
Capabilities
Context window262K tokens
Maximum output33K tokens
Openness Index44 / 100
OpenRouter parameters
frequency_penaltyinclude_reasoninglogit_biaslogprobsmax_tokensmin_ppresence_penaltyreasoningrepetition_penaltyresponse_formatseedstoptemperaturetool_choicetoolstop_ktop_logprobstop_p
Input
text
Output
text
Performance
Output speed302 tokens/s
Time to first token (TTFT)2.54 s
Time to first answer9.16 s
OpenRouter weekly rank#58
End-to-end response (500 t)10.8 s
Pricing
Input tokens$0.07
Cache Hit$0.010
Output tokens$0.22
Blended (3:1)$0.11